LTE Optimization Handbook TLA6.0 Document number: LTE/IRC/APP/032749 Document issue: V06.03 / EN Document status: Approved Standard Data classification: Date:
Confidential 28/Oct/2013
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
Copyright 2012 Alcatel-Lucent, All Rights Reserved UNCONTROLLED COPY: The master of this document is stored on an electronic database and is “write protected”; it may be altered only by authorized persons. While copies may be printed, it is not recommended. Viewing of the master electronically ensures access to the current issue. Any hardcopies taken must be regarded as uncontrolled copies. ALCATEL-LUCENT CONFIDENTIAL: The information contained in this document is the property of AlcatelLucent. Except as expressly authorized in writing by Alcatel-Lucent, the holder shall keep all information contained herein confidential, shall disclose the information only to its employees with a need to know, and shall protect the information from disclosure and dissemination to third parties. Except as expressly authorized in writing by Alcatel-Lucent, the holder is granted no rights to use the information contained herein. If you have received this document in error, please notify the sender and destroy it immediately. Without notice. Networks assumes no responsibility for errors that might appear in it .All other brand and product names are trademarks or registered are trademarks of their respective holders.
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
1 INTRODUCTION The document is the optimization handbook for the main RF features and related parameters per domain of Alcatel-Lucent LTE release TLA6.0 The procedures detailed in this document can be used to improve the network performance so that it meets contractual and technical objectives prior to a commercial launch. It can also be used in a continuous process as the network evolves due to addition of new cells, increase in traffic load or introduction of new features.
1.1 SCOPE Main purpose of the document consists of proposing parameter tuning that shall mitigate observed performance degradations. The following domains are distinguished for parameter tuning: Coverage Throughput Latency Capacity Mobility (both connected and idle mode) The parameters described in this document are related to the Alcatel-Lucent LTE TLA6.0 release.
1.2 AUDIENCE FOR THIS DOCUMENT The audience of this document is typically involved in following activities: Radio Network Design Radio Network Optimization First-Of-Arrial Trials Commercial Network Deployment
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 2/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
2 PUBLICATION HISTORY Version.Ed/ Language
Date (dd.mmm.yyyy)
06.00/EN
30.Jan.2013
06.01/EN
20.Feb.2013
06.02/EN
20.May.2013
06.03/EN
12.Oct.2013
Status
Name
MinMin Chen; Draft Yaguang Dong; Peng Xu MinMin Chen; Draft Yaguang Dong; Peng Xu MinMin Chen; Approved Yaguang preliminary Dong; Peng Xu
Approved Standard
Reason of changes / short description of significant changes to previous edition Creation
Updated according to TLA6.0.1
Updated according to TLA6.0.2 & Reading Cycle Completed
Added TM3/8 switching parameter Updated some description of MinMin transmission mode Chen; comparison Yaguang Updated default value of Dong; parameters according to the Peng Xu latest TLA6.0 MIM default template v34. Updated according to Reading Cycle Completed.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 3/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
TABLE OF CONTENTS
1 INTRODUCTION ............................................................................................................................... 2 1.1 SCOPE ........................................................................................................................................ 2 1.2 AUDIENCE FOR THIS DOCUMENT ......................................................................................................... 2 2 PUBLICATION HISTORY ................................................................................................................... 3 3 REFERENCE DOCUMENTS .............................................................................................................. 15 4 RELEASE RELATED DOCUMENTS ................................................................................................... 15 5 PROCESS AND METHODOLOGY ...................................................................................................... 16 5.1 TOOLS AND RESOURCES ................................................................................................................. 16 5.2 OPTIMIZATION TOOLS ................................................................................................................... 16 6 OPTIMIZATION PARAMETERS OVERVIEW....................................................................................... 17 6.1 PARAMETERS IMPACTING COVERAGE ................................................................................................. 18 6.2 PARAMETERS IMPACTING ACCESS ..................................................................................................... 19 6.3 PARAMETERS IMPACTING THROUGHPUT ............................................................................................. 21 6.4 PARAMETERS IMPACTING LATENCY ................................................................................................... 23 6.5 PARAMETERS IMPACTING CAPACITY .................................................................................................. 24 6.6 PARAMETERS IMPACTING MOBILITY .................................................................................................. 25 6.6.1 LTE – LTE Mobility .......................................................................................................... 25 6.6.2 LTE – UMTS Mobility ....................................................................................................... 27 6.6.3 LTE – GSM Mobility ......................................................................................................... 28 7 NEW FEATURES IN TLA6.0 ............................................................................................................ 29 8 COVERAGE OPTIMIZATION HINTS .................................................................................................. 31 8.1 PARAMETERS OPTIMIZATION FOR IMPROVING DOWNLINK COVERAGE .......................................................... 31 8.1.1 Referencesignalpower.................................................................................................... 31 8.1.2 PhichResource................................................................................................................ 33 8.1.3 n310 and t310 ................................................................................................................ 34 8.2 PARAMETERS OPTIMIZATION FOR IMPROVING UPLINK COVERAGE .............................................................. 35 8.2.1 sIRTargetforReferencePUCCHFormat ............................................................................. 35 8.2.2 sEcorrInit, sEcorrstepforlowerbler & secorrstepforhigherbler ........................................ 35 8.2.3 ulSyncSINRsyncToOOSTreshold & UlSyncSINROOStoSyncTreshold ................................... 37 8.2.4 deltaFPUCCHFormat1..................................................................................................... 39 8.2.5 eNB Tx Power Parameters Set ........................................................................................ 39 8.2.6 CELL COVERAGE ............................................................................................................. 46 8.2.6.1 UL CELL COVERAGE............................................................................ 46 8.2.6.2 DL CELL OUtdoor COVERAGE ................................................................ 47 8.2.6.3 TOTAL CELL COVERAGE....................................................................... 47 8.2.6.4 PUSCH FRACTIONAL POWER CONTROL ..................................................... 47 8.2.7 PUSCHPOWERCONTROLALPHAFACTOR............................................................................ 48 8.2.8 qRxLevMin ...................................................................................................................... 49 8.2.9 p0NominalPUSCH ........................................................................................................... 53 8.2.10 uplinkSIRtargetValueForDynamicPUSCHscheduling ....................................................... 55 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 4/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
8.3 FEATURE LINKED ......................................................................................................................... 57 8.3.1 T115590- Support of Multi-RRH Per Cell (One Logic Cell) for Indoor Coverage ............... 58 8.3.1.1 High Level Description and Benefits: ...................................................... 58 8.3.1.2 How to activate: ............................................................................... 59 8.3.1.3 Feature Impacts on eNB & Tunable Parameters: ........................................ 59 9 ACCESS OPTIMIZATION HINTS ....................................................................................................... 60 9.1 PARAMETERS OPTIMIZATION FOR IMPROVING ATTACH/DETACH PROCEDURES ............................................... 61 9.1.1 9.1.2 9.1.3 9.1.4 9.1.5
preambleInitialReceivedTargetPower ............................................................................ 61 PreambleTransmitPowerStepSize ................................................................................... 65 Scheduled TRANSMISSION (deltaPreambleMsg3 or tPCRACHMsg3)................................... 67 deltapreamblemsg3 ....................................................................................................... 70 tPCRACHMsg3 ................................................................................................................. 71
10 DOWNLINK THROUGHPUT OPTIMIZATION HINTS ........................................................................ 72 10.1 PARAMETERS OPTIMIZATION FOR IMPROVING DOWNLINK THROUGHPUT .................................................... 73 10.1.1 dlMCSTransitionTable ................................................................................................... 73 10.1.2 dlSinrThresholdBetweenCLMimoOneLayerAndTxDiv ..................................................... 76 10.1.3 dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer.............................................. 76 10.1.4 dlSinrThresholdBetweenOLMimoAndTxDiv .................................................................... 78 10.1.5 dlSINRThresholdbetweenRank1BeamformingAndTM3.................................................... 81 10.1.6 dlSINRThresholdbetweenRank2BeamformingAndTM3.................................................... 84 10.1.7 deltaSINRforIntermodeSwitch....................................................................................... 88 10.1.8 Downlink Performance Comparison Of Different Transmission Mode ............................ 89 10.1.8.1 Performance Comparison in CMCC LST DensityUrban Scenario ........................ 91 10.1.8.2 Performance Comparison in Jinqiao OTA SubUrban Scenario.......................... 94 10.1.8.3 Performance Comparison of Both Scenarios .............................................. 97 10.1.9 AlphaFairnessfactor.................................................................................................... 106 10.1.10 DynamicCFIEnabled .................................................................................................. 109 10.1.11 CFI............................................................................................................................ 110 10.1.12 cfi1allowed .............................................................................................................. 110 10.1.13 cfi2allowed .............................................................................................................. 111 10.1.14 cfi3allowed .............................................................................................................. 111 10.1.15 Measurement Gap..................................................................................................... 111 10.2 FEATURE LINKED ..................................................................................................................... 115 10.2.1 T115742- Rel9 DL MU-MIMO BF (TM8 Rank1) ............................................................... 115 10.2.1.1 High Level Description and Benefits: ..................................................... 115 10.2.1.2 How to activate: .............................................................................. 116 10.2.1.3 Feature Impacts on KPI’s & Tunable Parameters: ...................................... 116 11 UPLINK THROUGHPUT OPTIMIZATION HINTS............................................................................ 118 11.1 PARAMETERS OPTIMIZATION FOR IMPROVING UPLINK THROUGHPUT ....................................................... 118 11.1.1 uplinkSIRtargetValueForDynamicPUSCHscheduling ..................................................... 118 11.1.2 pUSCHPowerControlAlphaFactor ................................................................................ 121 11.1.2.1 puschpowercontrolalphafactor combination tests ...................................... 124 11.1.2.2 PUSCHPOWERCONTROLALPHAFACTOR COMBINATION TESTS (LIVE NETWORK) .... 132 11.1.3 ulSchedPropFairAlphaFactor ...................................................................................... 133 11.1.4 Measurement Gap ...................................................................................................... 134 12 LATENCY OPTIMIZATION HINTS ................................................................................................ 135 12.1 PARAMETERS OPTIMIZATION LATENCY ........................................................................................... 135 12.1.1 Test RECOMMENDATION AND results........................................................................... 137 12.1.1.1 Attach latency ................................................................................ 137 12.1.1.2 Attach Latency Results from CMCC LST................................................... 138 12.1.1.3 Quick reminder for isIntraFreqMobilityAllowed and aUGtriggerDelayforRACHmsg4 142 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 5/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.4 12.1.1.5 12.1.1.6 12.1.1.7 12.1.1.8 12.1.1.9
AUG procedure ................................................................................ 143 aUGtriggerDelayforRACHmsg4 .............................................................. 143 Idle to active latency ........................................................................ 144 Idle to Active Latency Results From Lab Test ........................................... 145 Idle to Active Latency Results From CMCC LST.......................................... 146 Quick reminder for isIntraFreqMobilityAllowed and aUGtriggerDelayforRACHmsg4 150 12.1.1.10 Reference of all the SoftWare used in the Platform .................................. 150 12.1.1.11 Detach latency ............................................................................... 151 12.1.1.12 Ping Latency results from CMCC LST ..................................................... 152 13 CAPACITY OPTIMIZATION HINTS ............................................................................................... 157 13.1 PARAMETERS OPTIMIZATION FOR IMPROVING CAPACITY ...................................................................... 157 13.1.1 Enb CAPACITY CONFIGURATIONS ................................................................................ 157 13.1.2 alphaFairnessFactor ................................................................................................... 158 13.1.3 UlSchedPropFairAlphaFactor ...................................................................................... 159 13.2 UL PHYSICAL CHANNELS CONFIGURATION AND CAPACITY ANALYSIS ...................................................... 161 13.2.1 SRS Configuration and Capacity Analysis .................................................................... 161 13.2.1.1 The Function of Sounding RS ............................................................... 161 13.2.1.2 SRS Configuration in Current Release ..................................................... 161 13.2.1.3 Analysis of SRS Configuration Impact on Capacity ...................................... 166 13.2.2 PUCCH Configuration and Capacity Analysis ............................................................... 167 13.2.2.1 The Function of PUCCH ...................................................................... 167 13.2.2.2 PUCCH Configuration in Current Release ................................................. 168 13.2.2.3 Parameters of pucch channel configuration IN TLA3.0/4.x/5.x/6.0 ................ 173 13.2.2.4 Analysis of PUCCH Configuration Impact on Capacity.................................. 175 13.3 TLA6.0 PUCCH&SRS CONFIGURATIONS AND RECOMMENDATIONS ...................................................... 176 14 INTRA LTE MOBILITY OPTIMIZATION HINTS .............................................................................. 179 14.1 NEIGHBOUR CONFIGURATION ...................................................................................................... 179 14.1.1 Intra Frequency Neighour Declaration ........................................................................ 181 14.1.2 Inter Frequency Neighbour Declaration ...................................................................... 182 14.2 MOBILITY PARAMETERS ............................................................................................................. 183 14.2.1 Mobility in RRC IDLE MODE ......................................................................................... 184 14.2.1.1 Cell Selection & Reselection................................................................ 184 14.2.1.2 QRXLEVMIN..................................................................................... 187 14.2.1.3 SINTRASEARCH ................................................................................ 188 14.2.1.4 SNONINTRASEARCH ........................................................................... 188 14.2.1.5 QHYST .......................................................................................... 189 14.2.1.6 QOFFSETCELL.................................................................................. 190 14.2.1.7 QRXLEVMINOFFSET ........................................................................... 190 14.2.1.8 threshServingLow ............................................................................. 191 14.2.1.9 threshXlow ..................................................................................... 192 14.2.1.10 threshXHigh ................................................................................... 193 14.2.1.11 TRESELECTIONEUTRAN ...................................................................... 194 14.2.1.12 TRESELECTIONEUTRASFMEDIUM ........................................................... 194 14.2.1.13 TRESELECTIONEUTRASFHIGH .............................................................. 195 14.2.1.14 tevaluation.................................................................................... 196 14.2.1.15 NCELLCHANGEHIGH.......................................................................... 196 14.2.1.16 NCELLChANGEMEDIUM ...................................................................... 197 14.2.1.17 QHYSTSFHiGH ................................................................................ 198 14.2.1.18 QHYSTSFMEDIUM ............................................................................. 198 14.2.2 Mobility in RRC Connected Mode ................................................................................ 199 14.2.2.1 FILTERCOEFFICIENTRSRP .................................................................... 200 14.2.2.2 HYSTERESIS .................................................................................... 202 14.2.2.3 TIMETOTRIGGER .............................................................................. 203 14.2.2.4 CELLINDIVIDUALOFFSET...................................................................... 203 14.2.2.5 EVENTA3OFFSET .............................................................................. 204 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 6/290
LTE Optimization Handbook TLA6.0 14.2.2.6 14.2.2.7 14.2.2.8 14.2.2.9 14.2.2.10 14.2.2.11 14.2.2.12 14.2.2.13 14.2.2.14
MGR/TIPS/NEA
OFFSETFREQ ................................................................................... 205 THRESHOLDEUTRARSRP ...................................................................... 205 THRESHOLD2EUTRARSRP .................................................................... 206 REPORTINTERVAL ............................................................................. 207 MAXREPORTCELLS............................................................................ 207 REPORTAMOUNT ............................................................................. 208 Call flow for Inter-eNB mobility, X2 HO – UE in RRC Connected .................... 209 Call flow for Inter-eNB mobility, S1 HO – UE in RRC Connected..................... 211 Intra LTE Handover Optimization Examples In Field Test ............................ 212
15 IRAT MOBLILTY OPTIMIZATION HINTS ...................................................................................... 225 15.1 INTER-RAT MOBILITY STRATEGY ................................................................................................. 225 15.2 LTE-UMTS OPTIMIZATION HINTS................................................................................................ 226 15.2.1 Inter- RAT Mobility in RRC Idle Mode.......................................................................... 227 15.2.1.1 qRxLevMin...................................................................................... 231 15.2.1.2 sNonIntraSearch............................................................................... 232 15.2.1.3 threshServingLow ............................................................................. 233 15.2.1.4 threshXLow .................................................................................... 234 15.2.1.5 tReselectionUtra .............................................................................. 235 15.2.1.6 TRESELECTIONUTRASFMEDIUM ............................................................. 236 15.2.1.7 TRESELECTIONUTRASFHIGH ................................................................. 237 15.2.1.8 NCELLCHANGEHIGH .......................................................................... 237 15.2.1.9 NCELLCHANGEMEDIUM ....................................................................... 238 15.2.1.10 QHYSTSFHiGH ................................................................................ 238 15.2.1.11 QHYSTSFMEDIUM ............................................................................. 239 15.2.2 Field Test Results Of Inter- RAT Mobility in RRC Idle Mode ........................................ 240 15.2.3 Inter-RAT Mobility in RRC Connected Mode ................................................................ 242 15.2.3.1 UE measurements needed for PS HO to UTRA-TDD ..................................... 243 15.2.3.2 Redirection To Utran......................................................................... 247 15.2.3.3 THRESHOLDEUTRARSRPB2 ................................................................... 252 15.2.3.4 THRESHOLDutrarscp .......................................................................... 252 15.2.3.5 OFFSETFREQUTRA ............................................................................ 253 15.2.3.6 FILTERCOEFFICIENTOFQUANTITYCONFIGUTRA .......................................... 253 15.2.3.7 HYSTERESIS .................................................................................... 254 15.2.3.8 TIMETOTRIGGER .............................................................................. 255 15.2.3.9 REPORTINTERVAL ............................................................................. 256 15.2.3.10 MAXREPORTCELLS............................................................................ 256 15.2.3.11 REPORTAMOUNT ............................................................................. 257 15.3 LTE-GSM MOBILITY OPTIMIZATION HINTS ..................................................................................... 258 15.3.1 IDLE MODE .................................................................................................................. 258 15.3.1.1 qRxLevMin...................................................................................... 261 15.3.1.2 SNONINTRASEARCH ........................................................................... 262 15.3.1.3 THRESHSERVINGLOW ......................................................................... 263 15.3.1.4 THRESHXLOW .................................................................................. 264 15.3.1.5 TRESELECTIONGERAN ........................................................................ 265 15.3.1.6 TRESELECTIONGERANSFMEDIUM ............................................................ 266 15.3.1.7 TRESELECTIONGERANSFHIGH ............................................................... 266 15.3.1.8 NCELLCHANGEHIGH .......................................................................... 267 15.3.1.9 NCELLCHANGEMEDIUM ....................................................................... 268 15.3.1.10 QHYSTSFHiGH ................................................................................ 268 15.3.1.11 QHYSTSFMEDIUM ............................................................................. 269 15.3.2 ACTIVE MODE ............................................................................................................ 270 15.3.2.1 Redirection to GERAN ........................................................................ 270 15.3.2.2 Cell Change Order with/without NACC to Geran ....................................... 272 15.3.2.3 THRESHOLDEUTRARSRPB2 ................................................................... 275 15.3.2.4 THRESHOLDGERAN............................................................................ 276 15.3.2.5 OFFSETFREQGERAN........................................................................... 277 15.3.2.6 FILTERCOEFFICIENTOFQUANTITYCONFIGGERAN ........................................ 277 15.3.2.7 HYSTERESIS .................................................................................... 278 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 7/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
15.3.2.8 TIMETOTRIGGER .............................................................................. 279 15.3.2.9 REPORTINTERVAL ............................................................................. 280 15.3.2.10 MAXREPORTCELLS............................................................................ 280 15.3.2.11 REPORTAMOUNT ............................................................................. 281 16 ABBREVIATIONS AND DEFINITIONS............................................................................................ 281 16.1 ABBREVIATIONS ....................................................................................................................... 281
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 8/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
LIST OF FIGURES Figure 8.1-1: CRS boosting Vs. non-boosting test route (Field Results JQHQ OTA–SubUrban) ..................... 33 Figure 8.1-2: SINR with CRS boosting Vs. without CRS boosting (Field Results JQHQ OTA–SubUrban) ......... 33 Figure 8.2-1: UE Drop Position Vs UL coverage (Field Results CMCC LST–Urban) ........................................ 38 Figure 8.2-2: First RRC connection reconfiguration for PUCCH/SRS Position Vs UL coverage (Field Results CMCC LST–Urban) .................................................................................................................... 39 Figure 8.2-3: Throughput for single UE vs. Path loss (Lab environment) .................................................... 48 Figure 8.2-4: Cell coverage (idle & connected) vs. qRxlevmin – 2600MHz (Field Results CMCC LSTUrban)................................................................................................................................................. 51 Figure 8.2-5: DL coverage for UE in idle Vs UE in active (Field Results CMCC LST-Urban) .......................... 52 Figure 8.2-6: qRxLevMin Selection ............................................................................................................. 52 Figure 8.2-7: Po_pusch_Nominal Impact .................................................................................................... 55 Figure 8.2-8: Slope - PuschPowerControl vs. uplinkSIRtargetValueForDynamicPUSCHscheduling................ 57 Figure 8.3-1: Typical wireless network example ........................................................................................ 58 Figure 8.3-2: Key challenges of wireless network - coverage and capacity ................................................ 58 Figure 8.3-3: One logical cell typical architecture ..................................................................................... 59 Figure 9.1-1: preambleTransMax vs. preambleInitialRceivedTargetPower vs. preambleTransmitPowerStepSize ........................................................................................................ 62 Figure 9.1-2: Po_preamble impact on UE Tx Power vs. PL(RA) (TRY1) ....................................................... 63 Figure 9.1-3: Po_preamble impact on UE Tx Power vs. PL (RA).................................................................. 63 Figure 9.1-4: preambleTransMax vs. preambleInitialRceivedTargetPower vs. preambleTransmitPowerStepSize (example of values) ........................................................................ 64 Figure 9.1-5: preamble TxPower vs. preambleInitialRceivedTargetPower Vs preamblePowerStepSize (Field Results CMCC LST-Urban) .......................................................................................................... 66 Figure 9.1-6: preamble Re-transfer Number vs. preambleInitialRceivedTargetPower Vs preamblePowerStepSize (Field Results CMCC LST-Urban).................................................................... 67 Figure 9.1-7: Near Cell RA Success Rate vs. preambleInitialRceivedTargetPower (Field Results CMCC LST-Urban) .......................................................................................................................................... 67 Figure 9.1-8: Parameters dependency and relations .................................................................................. 69 Figure 9.1-9: RA Success Rate Vs deltaPreambleMsg3 Vs tPCRACHMsg3 (Field Results CMCC LST-Urban) 71 Figure 9.1-10: PUSCH TxPower Vs deltaPreambleMsg3 Vs tPCRACHMsg3 (Field Results CMCC LSTUrban)................................................................................................................................................. 72 Figure 10.1-1: Radio link Quality vs. MCS Robustness vs. Throughput ........................................................ 73 Figure 10.1-2: Radio link Quality vs. dlMCSTransition Table vs. Throughput .............................................. 74 Figure 10.1-3: Dl Sinr Threshold Example .................................................................................................. 77 Figure 10.1-4: CL 2Layer-1Layer SNR Switch Threshold: 10 dB (purple) vs. 12 dB (blue) AWGN (Lab results VzW) ........................................................................................................................................ 78 Figure 10.1-5: CL 2Layer-1Layer SNR Switch Threshold: 10 dB (purple) vs. 12 dB (blue) EPA 5Hz, Medium Correlation (Lab Results VzW)................................................................................................ 78 Figure 10.1-6: Dl Sinr Threshold Example .................................................................................................. 79 Figure 10.1-7: dl SINR Threshold for TM switching driver test route (CMCC LST urban).............................. 80 Figure 10.1-8: OL 2Layer-TxDiv SINR Switch Threshold vs Phy average throuthput (CMCC LST Density urban) 81 Figure 10.1-9: dlSINRThreshold for TM3/7 Switch vs. PDCP average throughput (CMCC LST Density urban) 83 Figure 10.1-10: dlSINRThreshold for TM3/7 Switch vs. 2 codewords rate (CMCC LST Density urban) .......... 83 Figure 10.1-11: dlSINRThreshold for TM3/7 Switch vs. TM7 rate (CMCC LST Density urban) ....................... 84 Figure 10.1-12: Test road path for TM3/8 switching (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) ..................................................................................................................... 86 Figure 10.1-13: PDCP average throughput TM3 vs. TM8 (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) ..................................................................................................................... 86 Figure 10.1-14: PDCP average throughput TM3 vs. TM3/8 switching (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) .................................................................................................. 87 Figure 10.1-15: dlSINRthreshold for TM3/8 switching vs. PDCP average throughput (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) ........................................................................ 88 Figure 10.1-16: dlSINRthreshold for TM3/8 switching vs. PDCP average throughput (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) ........................................................................ 88 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 9/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-17: Test route path for TM performance comparison (site Lvkai in CMCC LST density urban) 90 Figure 10.1-18: Test route path for TM performance comparison (OTA in Jinqiao_Suburban) .................... 91 Figure 10.1-19: SINR vs. Avg. Throughput of TM2/3/7/8 (site Lvkai in CMCC LST density urban) ............... 92 Figure 10.1-20: SINR vs. MCS of TM2/3/7/8 (site Lvkai in CMCC LST density urban)................................... 92 Figure 10.1-21: SINR vs. BLER of TM2/3/7/8 (site Lvkai in CMCC LST density urban) ................................. 93 Figure 10.1-22: SINR vs. CQI_0 of TM2/3/7/8 (site Lvkai in CMCC LST density urban) ................................ 93 Figure 10.1-23: SINR vs. Avg. Throughput of TM2/3/7/8 (OTA in Jinqiao suburban) .................................. 95 Figure 10.1-24: SINR vs. MCS of TM2/3/7/8 (OTA in Jinqiao suburban) ...................................................... 95 Figure 10.1-25: SINR vs. BLER of TM2/3/7/8 (OTA in Jinqiao suburban) .................................................... 96 Figure 10.1-26: SINR vs. CQI_0 of TM2/3/7/8 (OTA in Jinqiao suburban) ................................................... 96 Figure 10.1-27: SINR vs. Avg. Throughput of TM3 in both scenarios ........................................................... 98 Figure 10.1-28: SINR vs. MCS of TM3 in both scenarios ............................................................................... 98 Figure 10.1-29: SINR vs. BLER of TM3 in both scenarios ............................................................................. 99 Figure 10.1-30: SINR vs. CQI_0 of TM3 in both scenarios ............................................................................ 99 Figure 10.1-31: SINR vs. Avg. Throughput of TM8 in both scenarios ......................................................... 100 Figure 10.1-32: SINR vs. MCS of TM8 in both scenarios ............................................................................. 100 Figure 10.1-33: SINR vs. BLER of TM8 in both scenarios ........................................................................... 101 Figure 10.1-34: SINR vs. CQI_0 of TM8 in both scenarios .......................................................................... 101 Figure 10.1-35: SINR vs. Avg. Throughput of TM2 in both scenarios ......................................................... 102 Figure 10.1-36: SINR vs. MCS of TM2 in both scenarios ............................................................................. 103 Figure 10.1-37: SINR vs. BLER of TM2 in both scenarios ........................................................................... 103 Figure 10.1-38: SINR vs. CQI_0 of TM2 in both scenarios .......................................................................... 104 Figure 10.1-39: SINR vs. Avg. Throughput of TM7 in both scenarios ......................................................... 104 Figure 10.1-40: SINR vs. MCS of TM7 in both scenarios ............................................................................. 105 Figure 10.1-41: SINR vs. BLER of TM7 in both scenarios ........................................................................... 105 Figure 10.1-42: SINR vs. CQI_0 of TM7 in both scenarios .......................................................................... 106 Figure 10.1-43: UEs’ distribution illustration ........................................................................................... 107 Figure 10.1-44: alphaFairnessFactor Change Impact with UE in different condition................................. 108 Figure 10.1-45: alphaFairnessFactor Change Impact – further details ...................................................... 109 Figure 10.1-46: Measurement Gap on DL and UL transmissions for TDD ................................................... 112 Figure 10.1-47: Test road path for MG comparison .................................................................................. 113 Figure 10.2-1: MU-MIMO Principle ............................................................................................................ 115 Figure 11.1-1: Impact of PUSCH Power for ul_PUSCH_SIRtarget (Field Results CMCC LST-Urban)............. 120 Figure 11.1-2: Impact of the Neighbor Cell IoT for ul_PUSCH_SIRtarget (Field Results CMCC LSTUrban)............................................................................................................................................... 121 Figure 11.1-3: Impact of the pUSCHPowerControlAlphaFactor =1.0 in MCS usage. ................................... 123 Figure 11.1-4: Impact of the pUSCHPowerControlAlphaFactor =0.8 in Throughput per RB ....................... 124 Figure 11.1-5: Impact of the pUSCHPowerControlAlphaFactor =0.7 in Throughput per RB. ...................... 124 Figure 11.1-6: Set 1 Result....................................................................................................................... 126 Figure 11.1-7: Set 1UL Throughput & UE TX Power vs. Path loss .............................................................. 127 Figure 11.1-8: SIR Target for theoretical assumptions with different alpha factor values ........................ 128 Figure 11.1-9: UL SIR Target for theoretical assumptions with different alpha factor values ................... 128 Figure 11.1-10: Different alpha factor comparison (Throughput, PRB’s, SINR & PUSCH SINR Target) ....... 129 Figure 11.1-11: UL Throughput & UE TX Power vs. Path loss alpha factor 0.7 & 1 with set 3 ................... 130 Figure 11.1-12: UL Throughput & UE TX Power vs. Path loss for alpha factor 0.7 for all sets................... 131 Figure 11.1-13: UL Throughput vs. Path loss for set1 & set3 with alpha factor 0.7 and 1 ........................ 132 Figure 11.1-14: UE TX Power vs. Path loss for set1 & set3 with alpha factor 0.7 and 1 ............................ 132 Figure 11.1-15: Impact of the ulSchedPropFairAlphaFactor ..................................................................... 133 Figure 12.1-1: Example for IMSI attach procedure (with authentication) ................................................. 137 Figure 12.1-2: Example for GUTI attach procedure (no authentication)................................................... 138 Figure 12.1-3: Initial Attach latency comparison overview ...................................................................... 139 Figure 12.1-4: Initial Attach latency and ratio (with authentication and Identity) in CMCC LST.............. 139 Figure 12.1-5: Initial attach latency example (without authentication and Identity) in CMCC LST ........... 140 Figure 12.1-6: Initial Attach procedure and latency example (with authentication and Identity) ............ 140 Figure 12.1-7: aUGtriggerDelayforRACHmsg4 Vs. Attach latency ............................................................. 140 Figure 12.1-8: isIntraFreqMobilityAllowed Vs. Attach latency .................................................................. 141 Figure 12.1-9: Authentication configuration in MME................................................................................. 142 Figure 12.1-10: AUG flow chart................................................................................................................ 143 Figure 12.1-11: Idle to active message chart ........................................................................................... 145 Figure 12.1-12: Example of total Idle to active latency ........................................................................... 146 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 10/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-13: Idle to active latency results from CMCC LST .................................................................. 147 Figure 12.1-14: aUGtriggerDelayforRACHmsg4 vs. latency ....................................................................... 148 Figure 12.1-15: isIntraFreqMobilityAllowed vs. latency ............................................................................ 148 Figure 12.1-16: SINR vs. latency .............................................................................................................. 149 Figure 12.1-17: Idle to activity message chart and latency ...................................................................... 149 Figure 12.1-18: Authentication configuration in MME ............................................................................... 150 Figure 12.1-19: No RF parameter optimization possible for detach latency! ............................................ 151 Figure 12.1-20: Non pre-scheduled and pre-scheduled Ping latencies with 32 byte payload (CMCC LST-Urban) ........................................................................................................................................ 152 Figure 12.1-21: Non-scheduled and pre-scheduled Ping latencies with 1400 byte payload (CMCC LSTUrban)............................................................................................................................................... 153 Figure 12.1-22: Overall Results (CMCC LST-Urban) ................................................................................... 154 Figure 12.1-23: Pre-scheduled vs. Non Scheduled Ping latency (32 Bytes) in Low SNR and High SNR (CMCC LST-Urban) ............................................................................................................................. 155 Figure 12.1-24: Pre-scheduled vs. Non Scheduled Ping latency (1432 Bytes) in Low SNR and High SNR (CMCC LST-Urban) ............................................................................................................................. 156 Figure 13.2-1: SoundingRS-UL-Config information element ...................................................................... 162 Figure 13.2-2: SRS frequency allocation example .................................................................................... 163 Figure 13.2-3: SRS capacity calculation example ..................................................................................... 166 Figure 13.2-4: PUCCH with P-CQI capacity calculation example .............................................................. 176 Figure 14.1-1: Coverage (Best Cell ID) analysis of South Pudong road area in CMCC LST – done by ARFCC team ...................................................................................................................................... 180 Figure 14.1-2: X2-Link access................................................................................................................... 181 Figure 14.1-3: LTE Intra Frequency Neighbour list .................................................................................. 182 Example below in Figure 14.1-4 shows the sector LST008_1 have Intra neighbours of sector LST008_2 and LST008_3, etc. ........................................................................................................................... 182 Figure 14.1-5: LTE Inter Frequency Neighbour list ................................................................................... 183 Figure 14.2-1: LTE to LTE Mobility – Measurement phase (RSRP vs. Time) ............................................... 185 Figure 14.2-2: LTE to LTE Mobility – Ranking Phase.................................................................................. 185 Figure 14.2-3: LTE to LTE Mobility – Decision Phase (RSRP vs. Time) ....................................................... 186 Figure 14.2-4: LTE to LTE Mobility – Measurement phase (RSRP vs. Time) ............................................... 187 Figure 14.2-5: LTE to LTE Mobility – Handover cases................................................................................ 199 Figure 14.2-6: Theoretical view ............................................................................................................... 201 Figure 14.2-7: filterCoefficientRSRP - Theoretical comparison (Simulation Analysis) ............................... 201 Figure 14.2-8: Call flow for Inter-eNB mobility, X2 HO – UE in RRC Connected (1) ................................... 209 Figure 14.2-9: Call flow for Inter-eNB mobility, X2 HO – UE in RRC Connected (2) ................................... 210 Figure 14.2-10: Call flow for Inter-eNB mobility, S1 HO – UE in RRC Connected (1) ................................. 211 Figure 14.2-11: Call flow for Inter-eNB mobility, S1 HO – UE in RRC Connected (2) ................................. 212 Figure 14.2-12: Test route and RSRP before optimization ........................................................................ 213 Figure 14.2-13: UE Signalling and Events in target cell before optimization ............................................ 213 Figure 14.2-14: Test route and RSRP before optimization ........................................................................ 214 Figure 14.2-15: UE Signalling and Events in target cell before optimization ............................................ 215 Figure 14.2-16: No measconfig received when UE attach in source cell before optimization ................... 215 Figure 14.2-17: RSRP became better after optimization .......................................................................... 216 Figure 14.2-18: UE successfully HO to target cell after optimization ....................................................... 216 Figure 14.2-19: Detail of measConfig message received after optimization ............................................. 217 Figure 14.2-20: Test route and RSRP before optimization ........................................................................ 218 Figure 14.2-21: UE Signalling and Events in source cell before optimization ............................................ 218 Figure 14.2-22: Detail signalling message when no measurement report before optimization ................. 219 Figure 14.2-23: RSRP became better after optimization .......................................................................... 220 Figure 14.2-24: source eNB received MR and UE successfully HO to target cell after optimization .......... 220 Figure 14.2-25: Detail of measConfig message received in source cell after optimization ....................... 221 Figure 14.2-26: Test route and RSRP before optimization ........................................................................ 222 Figure 14.2-27: UE Signalling and Events in source cell before optimization ............................................ 222 Figure 14.2-28: Detail signalling message of measConfig and received measurement report in source cell before optimization ................................................................................................................... 223 Figure 14.2-29: HO ping pong area between cell_121 and cell_108 (CMCC LST-Density urban) ................ 224 Figure 15.1-1: 2G/3G/LTE RAT selection example when multimode UE switch on ................................... 225 Figure 15.1-2: LTE TDD/Utran TDD RAT coverage gap example ............................................................... 226 Figure 15.2-1: LTE to UTRAN mobility in the context of IRAT mobility..................................................... 227 Figure 15.2-2: Cell Reselection procedure ............................................................................................... 228 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 11/290
LTE Optimization Handbook TLA6.0 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure
MGR/TIPS/NEA
15.2-3: UE rules follow-up............................................................................................................. 229 15.2-4: LTE to UTRAN Mobility – (RSRP vs. Time) measurement phase .......................................... 230 15.2-5: LTE to UTRAN Mobility – Algorithm Cell Reselection toward lower priority UTRAN Cell..... 231 15.2-6: LTE to UTRAN Mobility – (RSRP vs. Time) Decision Phase .................................................. 231 15.2-7: LTE TDD/Utran TDD IRAT cell reselection driver test route .............................................. 240 15.2-8: Detail signalling and events example of LTE TDD/Utran TDD IRAT cell reselection ........... 241 15.2-9: Events state machine ........................................................................................................ 243 15.2-10: UE measurements needed for PS HO to UTRA-TDD .......................................................... 243 15.2-11: Call flow for PS HO – Preparation phase .......................................................................... 244 15.2-12: Call flow for PS HO – Execution phase ............................................................................. 246 15.2-13: PS HO to UTRA-TDD - End-to-End call flows ................................................................... 247 15.2-14: LTE to UTRAN Mobility – Redirection Execution ............................................................... 247 15.2-15: RAT frequency with highest cellReselectionPriority is chosen for redirection ................. 248 15.2-16: RRC Connection Release with Redirection Info from EUTRAN to UTRAN .......................... 249 15.2-17: Inter RAT threshold for event B2 ..................................................................................... 249 15.2-18: Call flow for redirection from EUTRAN to UTRAN-Overview ............................................ 250 15.2-19: Call flow for redirection from EUTRAN to UTRAN-Description ......................................... 251 15.3-1: Reselection from eUTRAN to GERAN ................................................................................. 258 15.3-2: LTE to GERAN Mobility – HO to GERAN cell ........................................................................ 259 15.3-3: LTE to GERAN Mobility (RSRP vs. Time) – Cell Reselection – Measurement phase .............. 260 15.3-4: LTE to GERAN Mobility – Cell Reselection toward lower priority GERAN cell ..................... 260 15.3-5: LTE to GERAN Mobility (RSRP vs. Time) – Cell Reselection – Decision phase ...................... 261 15.3-6: Inter RAT threshold for event B2....................................................................................... 271 15.3-7: Call Flow for Redirection to Geran.................................................................................... 272 15.3-8: Serving Radio Condition and UE Measurement Configurations ........................................... 273 15.3-9: Call Flow for Cell Change Order with /Without NACC ....................................................... 275
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 12/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
LIST OF TABLES Table 6-1: Parameters impacting UL coverage ........................................................................................... 18 Table 6-2: Parameters impacting DL coverage ........................................................................................... 19 Table 6-3: Parameters impacting attach/detach procedures ..................................................................... 20 Table 6-4: Parameters impacting DL throughput........................................................................................ 22 Table 6-5: Parameters impacting UL throughput........................................................................................ 23 Table 6-6: Parameters impacting control plane latency............................................................................. 24 Table 6-7: Parameters impacting eNB Capacity ......................................................................................... 25 Table 6-8: Parameters impacting measurements for intra-LTE mobility .................................................... 26 Table 6-9: Parameters impacting measurements for inter-frequency (idle mode) ..................................... 26 Table 6-10: Parameters impacting measurements for inter-frequency (active mode)................................ 26 Table 6-11: Parameters impacting LTE – UMTS Inter-Frequency (Idle Mode) .............................................. 27 Table 6-12: Parameters impacting LTE – UMTS Inter-Frequency (Active Mode) .......................................... 28 Table 6-13: Parameters impacting LTE – GSM (Idle Mode) .......................................................................... 28 Table 6-14: Parameters impacting LTE – GSM (Active Mode) ...................................................................... 29 Table 7-1: TLA6.0 New Features ................................................................................................................ 30 Table 8-1: In the trial mode, the default setting for parameter referenceSignalPower ............................. 32 Table 8-2: The factor m i for TDD .............................................................................................................. 34 Table 8-3: 2×20W RRH Power Recommended Values in TM2/3/4 (10 MHz bandwidth) ............................... 40 Table 8-4: 2×20W RRH Power Recommended Values in TM2/3/4 (15 MHz bandwidth) ............................... 41 Table 8-5: 2×20W RRH Power Recommended Values in TM2/3/4 (20 MHz bandwidth) ............................... 41 Table 8-6: 8×5W RRH Power Recommended Values in TM2/3 (10 MHz bandwidth)..................................... 42 Table 8-7: 8×5W RRH Power Recommended Values in TM2/3 (15 MHz bandwidth)..................................... 42 Table 8-8: 8×5W RRH Power Recommended Values in TM2/3 (20 MHz bandwidth)..................................... 43 Table 8-9: 8×5W RRH Power Recommended Values in TM7 (10 MHz bandwidth) ........................................ 43 Table 8-10: 8×5W RRH Power Recommended Values in TM7 (15 MHz bandwidth) ...................................... 44 Table 8-11: 8×5W RRH Power Recommended Values in TM7 (20 MHz bandwidth) ...................................... 44 Table 8-12: 8×5W RRH Power Recommended Values in TM8 (10 MHz bandwidth) ...................................... 45 Table 8-13: 8×5W RRH Power Recommended Values in TM8 (15 MHz bandwidth) ...................................... 45 Table 8-14: 8×5W RRH Power Recommended Values in TM8 (20 MHz bandwidth) ...................................... 46 Table 8-15: UL 2.6GHz ............................................................................................................................... 47 Table 8-16: Path Loss & UL Cell Range in Dense Urban in Car (Field Results CMCC LST) ............................ 47 Table 8-17: DL Cell Range in Dense Urban in Car (Field Results CMCC LST) ............................................... 47 Table 8-18: Parameters to activate feature ............................................................................................... 59 Table 8-19: Tuneable parameters .............................................................................................................. 60 Table 9-1 K PUSCH for TDD configuration 0-6 ................................................................................................ 70 Table 10-1: Examples of threshold tuning for a 10MHz band (academic only, not applied in any trial /project). ........................................................................................................................................... 75 Table 10-2: Theory Assumption on CFI Tuning ......................................................................................... 110 Table 10-3: Measurement Gap Pattern configurations ............................................................................. 111 Table 10-4: MG impact on DL&UL throughput performance of UL/DL config2/7 (CMCC TDD LTE precommercial deployment Qingdao - Urban) ........................................................................................ 114 Table 10-5: HO delay test results w/wo MG enabled (CMCC TDD LTE pre-commercial deployment Qingdao - Urban)............................................................................................................................... 114 Table 10-6: Test SW configuration Reference .......................................................................................... 115 Table 10-7: Parameters to activate feature ............................................................................................. 116 Table 10-8: Tuneable parameters ............................................................................................................ 117 Table 11-1: uplinkSIRtargetValueForDynamicPUSCHscheduling vs. PUSCHPowerControlAlphaFactor........ 119 Table 11-2: Different Set’s Combinations ................................................................................................ 125 Table 12-1: SW Reference........................................................................................................................ 151 Table 12-2: Ping Latency for 32 Bytes with and without prescheduled for U-plane latency (CMCC LST-Density urban)............................................................................................................................ 152 Table 12-3: Test SW configuration Reference .......................................................................................... 156 Table 13-1: TLA6.0 Capacity figures ........................................................................................................ 158 Table 13-2: TLA6.0 SRS bandwidth configuration ..................................................................................... 164 Table 13-3: Power limitation of different SRS bandwidth configuration................................................... 165 Table 13-4: SRS subframe configuration................................................................................................... 165 Table 13-5: ISRS - UE Specific SRS Periodicity and Subframe Offset Configuration for TDD ........................ 165 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 13/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Table 13-6: Supported PUCCH formats..................................................................................................... 167 Table 13-7: PUCCH common parameters.................................................................................................. 169 Table 13-8: PUCCH dedicated parameters ............................................................................................... 169 Table 13-9: SR configuration parameters ................................................................................................. 170 Table 13-10: SR configuration Index ........................................................................................................ 170 Table 13-11: PUCCH CQI/PMI report configuration Index ......................................................................... 171 Table 13-12: PUCCH RI report configuration Index ................................................................................... 171 Table 13-13: SR configuration parameters ............................................................................................... 172 Table 13-14: PUCCH Format 1a/1b configuration parameters .................................................................. 172 Table 13-15: Per-UE Parameters from LUTs for TDD ................................................................................ 173 Table 13-16: Per-Cell Parameters from MIM and LUTs since TLA3.0 ......................................................... 173 Table 13-17: Additional Per-Cell and CQI related parameters to pre-calculate LUTs for TDD .................. 175 Table 13-18: Maximum number of users supported according to PUCCH formats ..................................... 175 Table 13-19: Uplink control channel/signal LookUpTable-Index0 ............................................................. 177 Table 13-20: Uplink control channel/signal LookUpTable-Index1 ............................................................. 178 Table 13-21: Uplink control channel/signal LookUpTable-Index2 ............................................................. 178 Table 14-1: Some Mobility parameters ..................................................................................................... 184 Table 15-1: IRAT cell reselection between LTE TDD & Utran TDD test results in CMCC LST (Density Urban)............................................................................................................................................... 241 Table 15-2: LTE TDD SW Reference.......................................................................................................... 242
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 14/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
3 REFERENCE DOCUMENTS Ref.
Document Number
Title
[1] [2]
LTE/IRC/APP/031548 LTE/IRC/APP/032749 MIM 15.1.3 template – W1339.5 – V34 LTE/SYS/APR/034643 v01.07 draft
LPUG TLA6.0 TLA5.1 Optimization Handbook
[3] [4]
TLA6.0.2-MIM15.1.3-template-v34.xtpl FTS for FRS T115590 Support of Multi-RRH per Cell (One Logic Cell) for Indoor Coverage
4 RELEASE RELATED DOCUMENTS Ref.
Document Number
Title
[1] [2] [3] [4] [5] [6] [7]
LTE/IRC/APP/032318 LTE/IRC/APP/031966 LTE/IRC/APP/031688 LTE/IRC/DJD/034923 LTE/IRC/APP/032278 LTE/IRC/APP/034221 LTE/IRC/APP/038290
LA5.0 Service Performance Handbook LA5.0 Migration - QoS and Stability Monitoring Handbook LA5.0 Engineering Toll Recommendation and Strategy LTE NEA TestPlan TLA6.0_V03.01 LA5.0 LTE Optimisation Troubleshooting Handbook LA5.0 LIMO Executive Report User Guide TLA6.0 Monitoring handbook KPI Design&FieldGuideline
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 15/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
5 PROCESS AND METHODOLOGY 5.1 TOOLS AND RESOURCES Tools used for the Network Optimization are described in the “LTE Engineering Tool Recommendation and Strategy” document.
5.2 OPTIMIZATION TOOLS A set of tools are needed to carry out the optimization objectives: Simulation Tool for RF design: Alcatel-Lucent uses A9155. Used for antenna change validation and RF analysis. There must be consistency with the methods used by the customer so it is possible to use their solution if it has been validated by Alcatel-Lucent teams. Data Acquisition Platform (DAP): currently Nixt Platform, composed of JSDU Nixt E6474A, W1314A receiver and 1 to 4 test mobiles. Tests Mobiles: typically Hisilicon, Qualcomm, Innofidei. Also IPW (Altair) or Sequans UEs. Post processing platform (PPP): CDS (Hugeland) which allow post-processing UE DT trace or Gladiator which allow automatic and customized KPI generation, built-in failure characterization, as well as UE and Call Trace synchronization capabilities (for a deeper and accurate analysis). For enhanced troubleshooting, the usage of a protocol analyzer (Agilent DNA, Nethawk, etc ) may be required, in particular to monitor the S1 and X2 interfaces. Project Database: For a correct follow-up of all the optimization activities it is mandatory to have a common and unique project information system (Project Database) which contains the following information: Site configuration (geographical coordinates, antenna height/azimuth/tilt) Site issues Updated operations plan Implemented/Planned Changes Metrics: performance KPI, outage times, workload, delays, escalated issues etc. Work Orders and implementation delays Reporting templates Equipment tracking List of raised ARs
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 16/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
6 OPTIMIZATION PARAMETERS OVERVIEW This chapter is intended to present various parameters existing in ALU RAN, grouped per main area of interest in trials and network deployments. Grouping was made such that it reflects various types of tests that are being performed during trials and KPIs tests that might as well be addressed during trials and network deployment. Parameters have been split by following domains: Coverage Attach /Detach Throughput Latency Capacity Mobility
Due to the wide scope of mobility, the parameters impacting mobility have been further divided in: LTE – LTE mobility LTE – UTRA TDD (TD-SCDMA) mobility LTE - GSM mobility
Inside each group, parameters are ordered by the most important and relevant for optimization activities. They should be optimized in case of strong constraints for performance (very demanding KPI, strong competition). To each parameter is associated a recommended value that can be obtained from [1] for parameters that have not been yet optimized in field activities. The parameters for which a different, optimized, value have been obtained in various field tests, have recommended values specified in the corresponding paragraphs along with some precisions about the conditions in which the optimized value have been obtained (cluster, load).
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 17/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
6.1 PARAMETERS IMPACTING COVERAGE The parameters of TLA6.0 release having an impact on coverage are presented in the table below. Most often the coverage is determined by UL link thus for optimizing coverage it is needed to optimize levels of UL power. Below one can find the parameters commented and reviewed for DL & UL Coverage. Object
Name
Recommended Value
ULPowerControlConf
pUSCHPowerControlAlphaFactor
0.8
ULPowerControlConf
p0NominalPUSCH
Check Recommendation
EnbRadioConf
uplinkSIRtargetValueForDynamicPUSCH scheduling
Check Recommendation
EnbRadioConf
sEcorrInit
0
EnbRadioConf
sEcorrStepForLowerBler
Check Recommendation
EnbRadioConf
sEcorrStepForHigherBler
Check Recommendation
EnbRadioConf
ulSyncSINRsyncToOOSTreshold
Check Recommendation
EnbRadioConf
ulSyncSINROOStoSyncTreshold
Check Recommendation
CellSelectionReselectionConf
qRxLevMin
Check Recommendation
ULPowerControlConf
deltaFPUCCHFormat1
deltaFm2
ULPowerControlConf
sIRTargetforReferencePUCCHFormat
0.0 [dB]
ULPowerControlConf
minSIRtargetForFractionalPowerCtrl
0.0 [dB]
ULPowerControlConf
maxSIRtargetForFractionalPowerCtrl
22.0 [dB]
ULPowerControlConf
pathLossNominal
Check Recommendation
ULPowerControlConf
p0NominalPUCCH
-100 [dBm]
Table 6-1: Parameters impacting UL coverage Here, parameters for PRACH & PUCCH power control are optional.
Object
Name
Recommended Value
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 18/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation Check Recommendation
CellSelectionReselectionConf
qRxLevMin
PowerOffsetConfiguration
referenceSignalPower
PowerOffsetConfiguration
primarySyncSignalPowerOffset
PowerOffsetConfiguration
secondarySyncSignalPowerOffset
PowerOffsetConfiguration
pBCHPowerOffset
PowerOffsetConfiguration
pDCCHPowerOffsetSymbol
PowerOffsetConfiguration
pCFICHPowerOffset
PowerOffsetConfiguration
pHICHPowerOffset
PowerOffsetConfiguration
pbOffsetPdsch
PowerOffsetConfiguration
paOffsetPdsch
LteCell
cellDLTotalPower
PowerOffsetConfiguration
phichResource
one
CellL1L2ControlChannelsConf
dlTargetSINRTableForPDCCH
[10.50,11.00,10.50,0.0 0,0.00,0.00,13.50,13.5 0,10.50,10.50,5.00,6.0 0,5.00,0.00,0.00,0.00, 7.00,7.00,5.00,5.00,1. 75,2.50,1.75,0.00,0.00 ,0.00,3.00,3.00,1.75,1. 75,-0.50,1.00,0.50,0.00,0.00,0.00,1. 50,1.50,-0.50,-0.50]
CellL1L2ControlChannelsConf
pdcchAggregationLevelForCRNTIGrantsInC ommonSearchSpace
8
CellL1L2ControlChannelsConf
pdcchAggregationLevelForUESearchSpace
4
UeTimers
n310
n20
UeTimers
t310
ms2000
Table 6-2: Parameters impacting DL coverage
6.2 PARAMETERS IMPACTING ACCESS
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 19/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Network performance can be evaluated by some measures implying attach and detach procedures (e.g. attach time, detach time, attach success rate). The parameters impacting attach, detach procedures are listed in the table below. Object
Name
Recommended Value
CellRachConf
preambleInitialReceivedTargetPower
dBm-94
CellRachConf
preambleTransmitPowerStepSize
dB4
ULPowerControlConf
deltaPreambleMsg3
12
CellRachConf
tPCRACHMsg3
4dB
CellRachConf
preambleTransMax
n8
LteCellTDD
spare4
3868311659 (bit#0~#15 value=49259)
CellRachConfTDD
maxHARQmsg3Tx
4
CellRachConf
maximumNumberOfDLTransmisionsRACH Message4
4
CellRachConf
rootSequenceIndex
Set according to prachConfigurationIndex , also the Network Planning
CellRachConf
zeroCorrelationZoneConfig
12
CellRachConfTDD
prachConfigurationIndex
CellRachConf
prachFrequencyOffset
CellRachConf
numberOfRAPreambles
56
CellRachConf
macContentionResolutionTimer
Sf64
CellRachConf
pRACHDetectFalseAlarmProb
0dot1
CellRachConfTDD
receptionOfMsg1Timer
30
UeTimers
n310
n20
UeTimers
t310
ms2000
Set according to PRACH format Set according to PRACH format, also the capacity
Table 6-3: Parameters impacting attach/detach procedures
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 20/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
6.3 PARAMETERS IMPACTING THROUGHPUT The parameters that can impact the throughput, both on UL and on DL, are listed in the table below. On DL the throughput is most influenced by the type of antenna system that is being used/selected wile on UL by the required quality of the received signal, which forces higher powers of PUSCH channel. Note that some parameters can increase the DL throughput while decreasing the UL throughput in the meantime due to the asymmetry of TDD LTE. These parameters are marked by a ‘*’. Moreover, Some TM7/TM8 related parameters used for 8 Antennas are listed as well, because TM7 can improve cell-edge users’ performance and TM8 can improve both near-cell and cell-edge users’ performance; IRC related parameter is also listed since it can improve the UL performance. Object
Name
Recommended Value
EnbRadioConf
dlMCSTransitionTable
Check table
DownlinkMimo DownlinkMimo DownlinkMimo
dlSinrThresholdBetweenCLMimoOneLaye rAndTxDiv dlSinrThresholdBetweenCLMimoTwoLay ersAndOneLayer dlSinrThresholdBetweenOLMimoAndTxDi v
-10.0 12.0 Check Recommendation Set according to customer strategy Set according to customer strategy
LteCellTDD*
subframeAssignment
LteCellTDD
specialSubframePatterns
CellL2DLConf
AlphaFairnessfactor
1.0
CellL1L2ControlChannelsConf
DynamicCFIEnabled
ture
CellL1L2ControlChannelsConf
cFI
3
CellL1L2ControlChannelsConf
cFI1Allowed
True
CellL1L2ControlChannelsConf
cFI2Allowed
True
CellL1L2ControlChannelsConf
cFI3Allowed
True
CellL1L2ControlChannelsConf
cFIThreshold1
2
CellL1L2ControlChannelsConf
cFIThreshold2
6
CellL1L2ControlChannelsConf
cFIIncreaseTimer
5
CellL2DLConfTDD
dlBasicSchedulingMode
PF
LteCellTDD
transmissionMode
Check Recommendation
UEAdaptiveBeamForming
beamFormingAlgo
COM-EBB
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 21/290
LTE Optimization Handbook TLA6.0
UEAdaptiveBeamForming UEAdaptiveBeamForming AdaptiveTransmissionModeSwit ch AdaptiveTransmissionModeSwit ch AdaptiveTransmissionModeSwit ch
MGR/TIPS/NEA
uLCESINRThresholdBetweenTxDivAndBe amFormingIntraTm7 sinrOffsetForBeamformingCQICompensa tion dlSINRThresholdbetweenRank1Beamfor mingAndTM3 dlSINRThresholdbetweenRank2Beamfor mingAndTM3
-17.0 3.0 Check Recommendation Check Recommendation
deltaSINRforIntermodeSwitch
3.0
UEAdaptiveBeamFormingTM8
beamFormingAlgoRank1
COM-EBB
UEAdaptiveBeamFormingTM8
beamFormingAlgoRank2
SU-BF-RANK2-COMEBB
UEAdaptiveBeamFormingTM8 UEAdaptiveBeamFormingTM8
dlSinrThresholdBetweenRank1BeamFor mingAndRank2BeamForming dlSinrThresholdBetweenTxDivAndRank1 BeamForming
0.0 0.0
UEAdaptiveBeamFormingTM8
rIThresholdBetweenRank1AndRank2
0.6
UEAdaptiveBeamFormingTM8
sinrOffsetForBeamformingPMICQI
0.0
UEAdaptiveBeamFormingTM8
sinrOffsetForBeamformingTxDivCQI
0.0
UEAdaptiveBeamFormingTM8
sinrOffsetForRank1AndRank2CW0
0.0
UEAdaptiveBeamFormingTM8
sinrOffsetForRank2CW0AndCW1
0.0
UEAdaptiveBeamFormingTM8 UEAdaptiveBeamFormingTM8 UEAdaptiveBeamFormingTM8 UEAdaptiveBeamFormingTM8
uLCESINRThresholdBetweenRank1BeamF ormingAndRank2BeamForming uLCESINRThresholdBetweenTxDivAndRa nk1BeamForming blerThresholdBetweenRank1BeamFormi ngAndRank2BeamForming blerThresholdBetweenTxDivAndRank1Be amForming
-51.2 -51.2 1.0 0.8
UEAdaptiveBeamFormingTM8
pmiRIReportR9
false
CellL1ULConfTDD
cqiReportingModeAperiodic
rm30
CellL1ULConfTDD
tddAckNackFeedbackMode
multiplexing
Table 6-4: Parameters impacting DL throughput Object
Name
Recommended Value
EnbRadioConf
uplinkSIRtargetValueForDynamicPUSCHs cheduling
Check Recommendation
ULPowerControlConf
pUSCHPowerControlAlphaFactor
0.8
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 22/290
LTE Optimization Handbook TLA6.0
CellL2ULConf
MGR/TIPS/NEA
ulSchedPropFairAlphaFactor
0.5
ulMCSTransitionTableForLargePUSCHGra nts ulMCSTransitionTableForSmallPUSCHGra nts
EnbRadioConf EnbRadioConf
Check table Check table
EnbRadioConfTDD
mCScorrectionForIRC
0.0
ULPowerControlConf
minSIRtargetForFractionalPowerCtrl
0.0 [dB]
ULPowerControlConf
maxSIRtargetForFractionalPowerCtrl
19.0 [dB]
ULPowerControlConf
pathLossNominal
Check Recommendation
ULPowerControlConf
p0NominalPUSCH
Check Recommendation
Table 6-5: Parameters impacting UL throughput
6.4 PARAMETERS IMPACTING LATENCY Latency is generally considered either as control plane latency or as user plane latency. Control plane latency involves the network attachment operation while user plane latency only considers the latency of packets while UE is in connected state. Parameters impacting control plane latency (attachment operation) and user plane latency are given in the tables below. Most of parameters impacting attachment operations are higher limits of various processes taking place during attachment procedure. The impact of such limits on the value of control plane latency is not significant. Object
Name
Recommended Value
CellRachConf
preambleInitialReceivedTargetPower
dBm-94
CellRachConf
preambleTransMax
n8
CellRachConf
preambleTransmitPowerStepSize
dB4
EnbRadioConf
aUGtriggerDelayforRACHmsg4
5
CellRachConf
macContentionResolutionTimer
Sf64
CellRachConfTDD
maxHARQmsg3Tx
4
CellRachConf
maximumNumberOfDLTransmisionsRACHMessage4
4
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 23/290
LTE Optimization Handbook TLA6.0
LteCellTDD
MGR/TIPS/NEA 3868311659 (bit#0~#15 value=49259)
spare4
Table 6-6: Parameters impacting control plane latency
6.5 PARAMETERS IMPACTING CAPACITY The parameters impacting capacity are listed in the table below. Object CellL2DLC onf CellL2ULC onf CellL1L2Co ntrolChann elsConf CellL1L2Co ntrolChann elsConf CellL1L2Co ntrolChann elsConf CellL1L2Co ntrolChann elsConf CellL1L2Co ntrolChann elsConf CellL2DLC onfTDD CellL2ULC onfTDD CellL2DLC onfTDD CellL2DLC onfTDD CellL2DLC onf CellL2DLC onf CellL2ULC onf
Name
Recommended Value
alphaFairnessFactor
1
ulSchedPropFairAlphaFa 0.5 ctor [10.50,11.00,10.50,0.00,0.00,0.00,13.50,13.50,10.50,10.50,5 dlTargetSINRTableForPD .00,6.00,5.00,0.00,0.00,0.00,7.00,7.00,5.00,5.00,1.75,2.50, CCH 1.75,0.00,0.00,0.00,3.00,3.00,1.75,1.75,-0.50,1.00,0.50,0.00,0.00,0.00,1.50,1.50,-0.50,-0.50] pdcchAggregationLevelF orCRNTIGrantsInCommo 8 nSearchSpace pdcchAggregationLevelF orNonCRNTIGrantsInCom 8 monSearchSpace pdcchAggregationLevelF 4 orUESearchSpace sINRThresholdBetweenA 30.0 L4andAL8 dlBasicSchedulingMode
PF
ulBasicSchedulingMode
PF
maxNumberOfRBsPerUE 100 maxGrantedUsers
9
maximumFSSUsers
32
maximumUsersInACQILis 32 tFromDLScheduler aperiodicCQIuserListMax 32 SizeInULS
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 24/290
LTE Optimization Handbook TLA6.0 CellL2ULC onf CellL1ULC onf
maxNbrULFSUsers
MGR/TIPS/NEA
32
uplinkControlChannelLU 0 Tindex
Table 6-7: Parameters impacting eNB Capacity
6.6 PARAMETERS IMPACTING MOBILITY Mobility in LTE includes mobility during idle states and during active states. Mobility can involve several technologies and several frequencies. Note: Measurement Gap feature parameters can impact the mobility parameters considered through the quality of measurement performed.
6.6.1 LTE – LTE MOBILITY Because measurements are somehow a common part of various types of mobility, in the table below are listed the parameters impacting measurement process for intra-LTE mobility. Object
Name
Recommended Value
CellSelectionReselectionConf
qRxLevMin
-140
CellSelectionReselectionConf
sIntraSearch
62
CellSelectionReselectionConf
qRxlevminoffset
2
CellSelectionReselectionConf
qHyst
dB2
CellReselectionConfLte
tReselectionEUTRAN
2
LteNeighboringCellRelation
qoffsetCell
dB0
RrcMeasurementConf
filterCoefficientRSRP
Fc8
LteSpeedDependentConf
tReselectionEutraSfMedium
oDot5
LteSpeedDependentConf
tReselectionEutraSfHigh
oDot25
SpeedStateEvalConf
tEvaluation
S30
SpeedStateEvalConf
nCellChangeHigh
12
SpeedStateEvalConf
nCellChangeMedium
4
SpeedStateEvalConf
qHystSfHigh
dB-6
SpeedStateEvalConf
qHystSfMedium
dB-6
ReportConfigEUTRA
Hysteresis
2
ReportConfigEUTRA
timeToTrigger
ms100
LteNeighboringCellRelation
cellIndividualOffset
dB0
ReportConfigEUTRA
eventA3Offset
2
LteNeighboringFreqConf
offSetFreq
dB0
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 25/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
ReportConfigEUTRA
reportInterval
ms240
ReportConfigEUTRA
maxReportCells
Check Recommendation
ReportConfigEUTRA
reportAmount
r8
Table 6-8: Parameters impacting measurements for intra-LTE mobility Object
Name
Recommended Value
CellSelectionReselectionConf
qRxLevMin
-140
CellSelectionReselectionConf
sNonIntraSearch
16
CellSelectionReselectionConf
threshServingLow
0
LteNeighboringCellRelation
threshXLow
0
CellSelectionReselectionConf
qHyst
dB2
CellReselectionConfLte
tReselectionEUTRAN
2
LteSpeedDependentConf
tReselectionEutraSfMedium
oDot5
LteSpeedDependentConf
tReselectionEutraSfHigh
oDot25
SpeedStateEvalConf
nCellChangeHigh
12
SpeedStateEvalConf
nCellChangeMedium
4
SpeedStateEvalConf
qHystSfHigh
dB-6
SpeedStateEvalConf
qHystSfMedium
dB-6
Table 6-9: Parameters impacting measurements for inter-frequency (idle mode) Object
Name
Recommended Value
ReportConfigEUTRA
thresholdEutraRsrp
-120
ReportConfigEUTRA
Threshold2EutraRsrp
-100
ReportConfigEUTRA
Hysteresis
2
ReportConfigEUTRA
timeToTrigger
ms100
RrcMeasurementConf
filterCoefficientRSRP
Fc8
LteNeighboringFreqConf
offSetFreq
dB0
ReportConfigEUTRA
reportInterval
ms240
ReportConfigEUTRA
maxReportCells
Check Recommendation
ReportConfigEUTRA
reportAmount
r8
Table 6-10: Parameters impacting measurements for inter-frequency (active mode)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 26/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
6.6.2 LTE – UMTS MOBILITY Coexistence of various technologies requires the possibility of performing mobility between various types of RAN. Indeed, such mobility requires multi-standard UEs. Parameters impacting LTE – UMTS mobility are presented in the table below. Object
Name
Recommended Value
CellReselectionConfUtraTdd
qRxLevMin
-115
CellSelectionReselectionConf
sNonIntraSearch
16
CellSelectionReselectionConf
threshServingLow
16
CellReselectionConfUtraTdd
threshXLow
0
UtraNeighboring
tReselectionUtra
2
UtraSpeedDependentConf
tReselectionUtraSfMedium
oDot5
UtraSpeedDependentConf
tReselectionUtraSfHigh
oDot25
SpeedStateEvalConf
nCellChangeHigh
12
SpeedStateEvalConf
nCellChangeMedium
4
SpeedStateEvalConf
qHystSfHigh
dB-6
SpeedStateEvalConf
qHystSfMedium
dB-6
Table 6-11: Parameters impacting LTE – UMTS Inter-Frequency (Idle Mode) Object
Name
Recommended Value
ReportConfigUTRA
thresholdEutraRsrpB2
-100
ReportConfigUTRA
thresholdUtraRscp
-114
RrcMeasurementConf
filterCoefficientOfQuantityConfigUtra
fc4
ReportConfigUTRA
hysteresis
4
ReportConfigUTRA
timeToTrigger
ms100
ReportConfigUTRA
reportInterval
ms240
ReportConfigUTRA
maxReportCells
1
ReportConfigUTRA
reportAmount
r8
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 27/290
LTE Optimization Handbook TLA6.0
MeasObjectUTRA
MGR/TIPS/NEA
offsetFreqUTRA
0
Table 6-12: Parameters impacting LTE – UMTS Inter-Frequency (Active Mode)
6.6.3 LTE – GSM MOBILITY Parameters impacting LTE – GSM mobility are presented in the table below. Object
Name
Recommended Value
CellReselectionConfGERAN
qRxLevMin
-101
CellSelectionReselectionConf
sNonIntraSearch
16
CellSelectionReselectionConf
threshServingLow
16
CellReselectionConfGERAN
threshXLow
0
GeranNeighboring
tReselectionGERAN
2
GeranSpeedDependentConf
tReselectionGERANSfMedium
oDot5
GeranSpeedDependentConf
tReselectionGERANSfHigh
oDot25
SpeedStateEvalConf
nCellChangeHigh
12
SpeedStateEvalConf
nCellChangeMedium
4
SpeedStateEvalConf
qHystSfHigh
dB-6
SpeedStateEvalConf
qHystSfMedium
dB-6
Table 6-13: Parameters impacting LTE – GSM (Idle Mode)
Object
Name
Recommended Value
ReportConfigGERAN
thresholdEutraRsrpB2
-100
ReportConfigGERAN
thresholdGeran
-110
RrcMeasurementConf
filterCoefficientOfQuantityConfigGERAN
fc2
ReportConfigGERAN
hysteresis
3
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 28/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
ReportConfigGERAN
timeToTrigger
ms100
ReportConfigGERAN
reportInterval
ms240
ReportConfigGERAN
maxReportCells
1
ReportConfigGERAN
reportAmount
r8
MeasObjectGERAN
offsetFreqGERAN
0
Table 6-14: Parameters impacting LTE – GSM (Active Mode)
7 NEW FEATURES IN TLA6.0 In the table below is presented all the new features belonging to TLA6.0 and it is identified all the domains impacted by each feature. For more information regarding the features the following link should be checked: TLA6.x FTS Documents for Review. Feature name DL dual-layer BF for 8 antenna (commercial) Rel9 DL MU-MIMO BF (TM8 Rank1) bCEM P1.1 for 8A configurations BBU Configurations for T/LA6.0 BBU configuration clarification for one logic cell (churn 166011) support of multiRRH per cell (one logic cell) for indoor coverage Band 38 (2.6GHz) LTE eNodeB Configurations in TLA6.0 Band 40 (2.3GHz) LTE eNodeB Configurations in TLA6.0
Optimiz ation
Capacity
Throughput
Coverage
Latenc y
Optim
x
x
x
x
Optim
x
Optim
x
x
x
HW/Opt im
x
x
HW/Opt im
x
E2E/Opt im
Mobilit y
Attac h
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
E2E
x
x
x
x
x
x
E2E
x
x
x
x
x
x
MO S
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 29/290
LTE Optimization Handbook TLA6.0
Feature name 3GPP 36.104/141 Conformance test in TDD TLA6.0 Release Upgrade Support support ZUC only (w/o SNOW3G) for 8A field trial TLA6.0 SW capacity target ATCR TD-RRH8x1026 (8x10W RRH) TLA6.0 8A L1 counter introduction
Optimiz ation
Capacity
RF/Evol
MGR/TIPS/NEA
Throughput
Coverage
Latenc y
Mobilit y
Attac h
x
x
x
x
x
x
x
MO S
KPI/OM C E2E E2E/Opt im RF/Opti m
x
x x
x
KPI/Opti m
Table 7-1: TLA6.0 New Features
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 30/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
8 COVERAGE OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is coverage in LTE. In this chapter it will be beaked in two sub-domains; Downlink Coverage and Uplink Coverage. Normally some questions arise, such as: When to perform coverage optimization? Which parameters can help extending /reducing the coverage?
Mainly the Coverage optimization can occur when the Link Budget is below expectations (Theoretical calculation). Before starting playing with the parameterization; usually is part of best practice rules for in Near Cell /Mid Cell & Cell Edge test to follow up simple steps as: Check the CQI Evaluate RSSI vs. SNR relation Evaluate RSRP vs. RSRQ Throughput Values for specific location
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If regardless of the correct values, still facing some performance issues, the below parameters can be used in order to correct the situation. When changing parameters; you can adopt a more error-free approach, meaning that a parameter is changed at each time. If three or four parameters are changed same time… it could be difficult to understand which one is bringing the improvement in performance. As note; please remember that this can be a static test in each position, or can be a moving test… the same principles can be applied in both situations.
8.1 PARAMETERS OPTIMIZATION FOR IMPROVING DOWNLINK COVERAGE 8.1.1 REFERENCESIGNALPOWER The Reference Signal Power is a key RF parameter that impacts coverage. Parameter referenceSignalPower configures the DL RS absolute power applied per Resource Element (REG) and per transmit antenna. This level is used as a power level reference (the power levels for all the other DL signals and channels are set relative to it). ATTENTION! When modifying this parameter, all other signal power setting will be adjusted in accordance to a re-calculated power offset relative to the referenceSignalPower. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 31/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
This parameter is expressed in dBm. It is converted into linear scale (miliwatts) according to the following formula: P [mW] = 100.1×referenceSignalPower Expected behaviour when changing this parameter: The higher the setting, the larger the cell coverage on the downlink, but leaves smaller power headroom available for other downlink signals and channels. The lower the value, the smaller the cell coverage on the downlink axis, but larger power headroom is available for other downlink signals and channels. Note: The following Table 8-1 translates the expected behaviour in terms of cellDLTotalPower when changing the reference Signal Power; some difference may occur if other sets of parameters are used also. TLA6.0 supports 8 antennas and 2 antennas with OLC eNB (One Logical Cell, also named Supper Cell), 2 Antennas OLC eNB supports transmission mode 1/2/3/4 and 8 Antennas eNB supports transmission mode 1/2/3/7/8. For RSRP, RSSI, RSRQ relationship, please refer to 3GPP TS36.214. Recommended & Default Value = Check Table 8-1 Antennas
transmissionMode
cellDLTotalPower (dBm) 43
Bandwidth
referenceSignalPower (dBm) 2 antennas TM1, TM2, TM3, n50-10MHz 18 TM4 (2×20w RRH) n100-20MHz 15 8 antennas TM1, TM2, TM3, 37 n50-10MHz 16 TM7, TM8 (8×5w RRH) n100-20MHz 13 8 antennas TM1, TM2, TM3, 40 n50-10MHz 16 (per of 2 carriers) or TM7, TM8 19 (single carrier) (8×10w RRH) n100-20MHz 13 (per of 2 carriers) or 16 (single carrier) Table 8-1: In the trial mode, the default setting for parameter referenceSignalPower Note that, since TLA6.0 MIM15.1.3 template v12, the CRS power boosting by increasing referenceSignalPower and decreasing paOffset are recommended (for detail info please refer to section 8.2.5). The recent change in terms of recommendation for the CRS booting is due to live testing at JQHQ OTA field test results, SINR increasing were seen after tuning power setting from the non-CRS power boosting values. The testing procedure should comprise the following steps: Step 1: Set the referenceSignalPower and other DL signal/channels power setting as default sets of values without RS boosting. Step 2: Connect the UE in Near-Cell radio conditions with DL full buffer FTP transfer and perform driver test towards Celledge till UE drop. For a more consistent data we recommend a drive back as well logged in another trace. Step 3: Using the same cell and same route, choose referenceSignalPower and other DL signal/channels as the sets of values with RS boosting and repeat Step2. Step 4: Post process the logged data and provide results in terms of RSRP, SINR and coverage statistic.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 32/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
JQHQ OTA RSRP Vs. SINR
34 32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0
Avg. SINR without CRS boosting Avg. SINR with CRS boosting
-123 -120 -117 -114 -111 -108 -105 -102 -99 -96 -93 -90 -87 -84 -81 -78 -75 -72 -69 -66
SINR
Figure 8.1-1: CRS boosting Vs. non-boosting test route (Field Results JQHQ OTA–SubUrban)
RSRP Figure 8.1-2: SINR with CRS boosting Vs. without CRS boosting (Field Results JQHQ OTA– SubUrban) Based on the test results, we can conclude that, SINR with RS boosting is about 2~5dB higher than that without RS boosting. This will make benefit for downlink channel quality estimation and improve DL coverage, although based on JQHQ OTA field test, basically no gain on throughput could be seen after CRS power boosting, further field tests will be done in CMCC commercial deployment.
8.1.2 PHICHRESOURCE PHICH channels are grouped in PHICH groups. Each PHICH group consists of 8 PHICH channels (hence conveys 8 ACK/NACKs) that use the same resources, PHICH channels of a same group being separated by orthogonal sequences. The number of PHICH groups in TDD subframe i is:
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 33/290
LTE Optimization Handbook TLA6.0
N
group PHI CH
MGR/TIPS/NEA
DL N g N RB 8 DL 2 N g N RB 8
for normal cyclic prefix for extended cyclic prefix
Where: Ng ∈{1/6 ,1/2 , 1, 2} and is configured by parameter phichResource NDLRB is the total number of RBs in the downlink and is configured by parameter FrequencyAndBandwidthTDD :: Bandwidth In TDD, the number of PHICH groups may vary between downlink subframes and is given by group , where m i is given below: mi N PHICH Uplink-downlink configuration 0 1 2 3 4 5 6
0 2 0 0 1 0 0 1
1 1 1 0 0 0 0 1
Subframe number 2 3 4 5 6 - - - 2 1 - - 1 0 1 - 1 0 0 0 - - - 0 0 - - 0 0 0 - 0 0 0 0 - - - 1 1
i 7 0 0 0 -
8 1 1 1 1 -
9 1 0 1 1 0 1
Table 8-2: The factor m i for TDD A PHICH group consists of 3 REGs over either 1 or 3 OFDM symbols, depending on the value of parameter phich-Duration (“normal” or “extended”). This parameter can only be set to “extended” if the CFI is equal to 3. Recommended Value & Only Supported Value= "1" Expected behaviour when changing this parameter: Setting the value low will result in lower number of PHICH groups in a subframe, so the higher the number of ACK/NACKs that need to be sent out the longer the buffer, eventually leading to failing to transmit the messages. Setting value high will impact in having a higher number of PHICH groups in a subframe, so the fewer ACK/NACKs needed to be transmitted, OFDM symbols are not used and the allocated resources for this process go to waste. Note that, currently, ALU LTE eNB only support Ng=1, i.e. parameter phichResource=1.
8.1.3 N310 AND T310 n310 defines the maximum number of consecutive "out-of-sync" indications received from lower layers for the UE to detect physical layer problems. It is broadcasted in SIB2. t310 specifies the start value for the UE timer T310. This timer is started in the UE in RRC connected mode upon detecting radio link problems. At timer expiry the UE will go to RRC idle mode if security is not activated, else initiate the RRC connection re-establishment procedure. It is broadcasted in SIB2. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 34/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
After UE low layer report n310 times of "DL out-of sync", UE high layer will start t310 timer and wait for UE low layer report "DL in-sync", if UE doesn't receive "DL in-sync" before t310 expired, then UE will assume it is "out of sync" in downlink.
Recommended & Default Value: n310= "n20", t310= "ms2000" Expected behaviour when changing these two parameters: Setting the value low will result in higher risk of OOS and easier to go to RRC idle (drop) or RRC reestablishment. Setting the value high will help to reduce the risk of OOS, thus the traffic transmission could be more stable, but may also cause UE hard to drop or re-establish RRC even when the RF condition is really poor.
8.2 PARAMETERS OPTIMIZATION FOR IMPROVING UPLINK COVERAGE 8.2.1 SIRTARGETFORREFERENCEPUCCHFORMAT The PUCCH power control procedure is used to guarantee the required error rate. For this purpose, it aims at achieving a target SIR the value of which guarantees the required error rate. The SIR target is set to sIRTargetforReferencePUCCHFormat for PUCCH Format 1A and to sIRTargetforReferencePUCCHFormat + deltaFPUCCHFormat1b for PUCCH format 1B. Note that the PUCCH power control procedure assumes shortened PUCCH Format to account for the SRS configuration. This parameter is a key RF optimization parameter. Higher settings of this parameter will improve PUCCH reception, but will also drive higher UE TX power leading to interference to neighbouring cells, and vice-versa.
Recommended & Default Value= “0.0”
8.2.2 SECORRINIT, SECORRSTEPFORLOWERBLER & SECORRSTEPFORHIGHERBLER The eNB starts Spectrum Efficiency Correction when the call setup is completed using these two parameters and sEcorrMin, sEcorrMax. sEcorrInit is the initial correction factor value applied at call setup.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 35/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
This parameter controls the initial value (in dB) of correction metric SEcorr managed by the link adaptation function to maintain the PUSCH initial HARQ BLER around its target value. The lower the sEcorrInit value, the more conservative is the PUSCH link adaptation starting point and, consequently, the lower the PUSCH MCS. In other words, the lower the sEcorrInit values, the lower is the initial PUSCH MCS value, and the lower are the risks of observing high BLER value at call setup or on the target cell just after handover. Note however that the lower the sEcorrInit value, the longer the link adaptation will take to converge to its setpoint. This may impact the maximum achievable throughput for a brief period (exact time depends on traffic activity) after call setup or handover. Recommended & Default Value= “0.0”, “-10.0” and “10.0” for sEcorrInit, sEcorrMin and sEcorrMax. Upon successful decoding of a HARQ process corresponding to an UL dynamic grant after N HARQ transmissions the spectrum efficiency correction factor for UE k shall be updated as follows:
SECorrn ew( k) SECorrold ( k) SEcorr _ step( N) sEcorrStepForLowerBLER and sEcorrStepForHigherBLER parameters are key RF optimization parameters regarding SEcorr _ step( N) . Higher step sizes will allow the eNodeB to make quicker compensation to any sources of SINR-to-MCS conversion error, but could lead to wider swings in data rates and HARQ performance. Lower settings will slow down the spectrum efficiency correction process, impacting HARQ performance. The tuning of the table of SEcorr_step values (i.e. sEcorrStepForLowerBLER and sEcorrStepForLowerBLER) is performed so that if N = floor(x) and N≠x, then the ratio of the SEcorr_step values around the target HARQ Tx rate satisfies the following condition:
SEcorr _ step[ N 1] ( x N ) 1 SEcorr _ step[ N ] xN Where • x target HARQ Tx rate is the target HARQ Tx rate and x>1. • N floor(x ) is the N HARQ transmission. The other values in the SEcorr_step tables are set in order to allow fast convergence around that set point. Recommended & Default Value= “[-0.50000000, 0.00625000, -0.06250000, -0.25000000, 0.50000000, -0.50000000, -0.50000000, -0.50000000, -0.50000000, -0.50000000]” for sEcorrStepForLowerBLER
& Recommended & Default Value= “[-0.50000000, 0.00625000, -0.06250000, -0.25000000, 0.50000000, -0.50000000, -0.50000000, -0.50000000, -0.50000000, -0.50000000]” for sEcorrStepForHigherBLER Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 36/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Note: The default value of parameter “sEcorrStepForHigherBLER” is set to same as parameter “sEcorrStepForLowerBLER” in TLA6.0 as well. Typical settings of sEcorrStepForLowerBLER provide around 10% HARQ 1st HARQ reTx performance.
8.2.3 ULSYNCSINRSYNCTOOOSTRESHOLD & ULSYNCSINROOSTOSYNCTRESHOLD The uplink synchronization detection mechanism is based on the SIR metric derived from the Sounding Reference Signal observations. Upon processing of an SRS measurement report from L1 related to user k, the UL scheduler evaluates the UL synchronization status of that user by comparing the SRS synchronization metric computed to threshold levels as follows: The threshold for the transition from “In sync state” to “out of sync state” is configured by parameter ulSyncSINRsyncToOOSThreshold, i.e. If the UE is assumed in “In Sync state” and the following condition is met: SINRsync(userk)
ulSyncSINROOStoSyncThreshold Then the user is considered in “In Sync state” by the MAC scheduler. Note that the tuning must satisfy the following condition: ulSyncSINRsyncToOOSThreshold < ulSyncSINROOStoSyncThreshold
Default Values = “-11.0” for ulSyncSINRsyncToOOSThreshold & “-10.0” for ulSyncSINROOStoSyncThreshold
NEA Recommended Value= “-17.0” for ulSyncSINRsyncToOOSThreshold & “-16.0” for ulSyncSINROOStoSyncThreshold
The recent change in terms of recommendation for the above parameters is due to live testing at CMCC LST Cluster were good results were seen after tuning them from the “official” values. The testing procedure should comprise the following steps: Step 1: Set the ulSyncSINROOStoSyncTreshold to -4 ulSyncSINRsyncToOOSThreshold to -5, then connect the UE in Near-Cell radio conditions with 128Kbps UL UDP transfer and perform driver test towards Celledge till eNB first time reset PUCCH/SRS resource for UE. Step 2: Continue performing driver test towards Celledge till UE drop. For a more consistent data we recommend a drive back as well logged in another trace.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 37/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 3: Using the same cell and same route choose another value for ulSyncSINROOStoSyncTreshold and ulSyncSINRsyncToOOSThreshold = {(-6,-7), (-8,-9), (-10,-11), (-12,-13), (-14,-15), (-16,-17), (-19,-20) } and repeat Step 1 and Step 2. Step 4: Post process the logged data and provide results in terms of First time RRC connection reconfiguration for PUCCH/SRS Position, UE Drop Position, UL Throughput and PUSCH BLER.
Using the test procedures above, the follow results are expected: The UE droped position and the first time eNB reset PUCCH/SRS resource for UE are increased as these two threshold decreasing.
UL Coverage (m)
Ue Drop Position 1420 1400 1380 1360 1340 1320 1300 1280 1260 1240 1220
UL Coverage (m)
ulSyncSINROOStoSyncTreshold , ulSyncSINRsyncToOOSTreshold Figure 8.2-1: UE Drop Position Vs UL coverage (Field Results CMCC LST–Urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 38/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
First RRC reconfiguration for PUCCH/SRS Position 1400 UL Coverage (m)
1350 1300 1250 1200 1150
UL Coverage (m)
1100 1050
ulSyncSINROOStoSyncTreshold , ulSyncSINRsyncToOOSTreshold Figure 8.2-2: First RRC connection reconfiguration for PUCCH/SRS Position Vs UL coverage (Field Results CMCC LST–Urban)
8.2.4 DELTAFPUCCHFORMAT1 This parameter is used for setting the transmit power of SR over PUCCH by the UE. TLA6.0 eNodeB relies on Scheduling Request on PUCCH from UE for scheduling uplink grants. If the SR is not received by the eNodeB, it will trigger the UE to declare SRmax failure, which in turn will trigger eNodeB’s OOS condition. Under certain conditions the SR power may not be sufficient to ensure detection. By boosting the SR transmit power, it increases the SR detection likelihood at the eNodeB. Concerning the deltaFPUCCHFormat1; the setting of the UE Transmit Power PPUCCH for PUCCH in
{
}
subframe i is defined by PPUCCH (i) = min Pmax ,P0 _ PUCCH + PL + ΔF _ PUCCH + g(i) [dBm] Where ΔF _ PUCCH denotes the (PUCCH) format specific power offset; the format dependent power offset ΔTF_PUCCH(TF) is defined on a per cell basis and configured by parameter deltaFPUCCHFormat1 for PUCCH format 1 relative to PUCCH format 1a. Recommended & Default Value= deltaFPUCCHFormat1= “deltaFm2”
8.2.5 ENB TX POWER PARAMETERS SET For coverage optimization process the primary set of parameters are for power settings which can have a big influence on coverage. The main coverage parameters are: referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 39/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The higher the power the better the coverage but also the interference in the neighbouring cells. The values must be tuned so that cells coverage with same frequency does not overlap. TLA6.0 supports 8 antennas and 2 antennas with OLC eNB (One Logical Cell, also named Supper Cell), 2 Antennas OLC eNB supports transmission mode 1/2/3/4 and 8 Antennas eNB supports transmission mode 1/2/3/7/8. Two transmission schemes are inside TM7, i.e., two ports TxDiv and single port5 BeamForming are used. Three transmission schemes are inside TM-8, i.e., two ports TxDiv , port7 or port8 single layer BeamForming and port7 & port8 dual-layer BeamForming are used. For BF scheme in TM7 and TM8, the PBCH, PCFICH, PDCCH & PHICH are transmitted using TxDIV (port 0 and 1). Due to TxDiv encoding, the PBCH, PCFICH, PDCCH & PHICH are transmitted at -3dB compared to the configured setting. The PDSCH uses UE-adaptive beam and other channels use broadcast beam. For TxDiv scheme in TM7 and TM8, due to TxDiv encoding, the PBCH, PCFICH, PDCCH & PHICH are transmitted at -3dB compared to the configured setting. All channels, including PDSCH and other channels, use broadcast beam. Note that power configuration is set per antenna port if there is no specific description in TDD. Due to the fixed maximum power of the PA, the maximum power per RE per antenna is 3dB less in 20MHz where the number of RE is doubled compared to 10 MHz and the maximum value for referenceSignalPower is lower in 20MHz bandwidth in TDD. The following tables are provided respectively for 20MHz, 15MHz & 10MHz bandwidths. All power offsets in the following tables are per antenna port. Note that the OAM value of the offsets marked with an asterisk (*) is for 2 antennas (ports), and is 3 dB greater than the value in the following tables (for example, -3.0 dB in table is coded as 0.0 dB in OAM). Recommended values for a 2×20W RRH Power & 10MHz in TM2/3/4: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 2×20W radio 18 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 43 dBm (20W)
Table 8-3: 2×20W RRH Power Recommended Values in TM2/3/4 (10 MHz bandwidth) Recommended values for a 2×20W RRH Power & 15MHz in TM2/3/4: Inputs
Recommended values 2×20W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 40/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 2×20W radio 16 dBm 0.2 dB 0.2 dB -3.0 dB * -2.6 dB * -2.6 dB * -2.0 dB * -2.0 dB * -3 dB 0 dB (PB = 1) 43 dBm (20W)
Table 8-4: 2×20W RRH Power Recommended Values in TM2/3/4 (15 MHz bandwidth) Recommended values for a 2×20W RRH Power & 20MHz in TM2/3/4: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 2×20W radio 15 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 43 dBm (20W)
Table 8-5: 2×20W RRH Power Recommended Values in TM2/3/4 (20 MHz bandwidth) Recommended values for a 8×5W RRH Power & 10MHz in TM2/3: Inputs
Recommended values 8×5W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 41/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 8×5W radio 16 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 37 dBm (5W)
Table 8-6: 8×5W RRH Power Recommended Values in TM2/3 (10 MHz bandwidth) Recommended values for a 8×5W RRH Power & 15MHz in TM2/3: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 8×5W radio 14 dBm 0.2 dB 0.2 dB -3.0 dB * -2.6 dB * -2.6 dB * -2.0 dB * -2.0 dB * -3 dB 0 dB (PB = 1) 37 dBm (5W)
Table 8-7: 8×5W RRH Power Recommended Values in TM2/3 (15 MHz bandwidth) Recommended values for a 8×5W RRH Power & 20MHz in TM2/3: Inputs
Recommended values 8×5W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 42/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch cellDLTotalPower
Recommended values 8×5W radio 13 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 37 dBm (5W)
Table 8-8: 8×5W RRH Power Recommended Values in TM2/3 (20 MHz bandwidth) Recommended values for a 8×5W RRH Power & 10MHz in TM7: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port5PowerOffset cellDLTotalPower
Recommended values 8×5W radio 16 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-9: 8×5W RRH Power Recommended Values in TM7 (10 MHz bandwidth) Recommended values for a 8×5W RRH Power & 15MHz in TM7: Inputs
Recommended values 8×5W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 43/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port5PowerOffset cellDLTotalPower
Recommended values 8×5W radio 14 dBm 0.2 dB 0.2 dB -3.0 dB * -2.6 dB * -2.6 dB * -2.0 dB * -2.0 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-10: 8×5W RRH Power Recommended Values in TM7 (15 MHz bandwidth) Recommended values for a 8×5W RRH Power & 20MHz in TM7: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port5PowerOffset cellDLTotalPower
Recommended values 8×5W radio 13 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-11: 8×5W RRH Power Recommended Values in TM7 (20 MHz bandwidth) Recommended values for a 8×5W RRH Power & 10MHz in TM8: Inputs
Recommended values 8×5W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 44/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port7port8PowerOffset cellDLTotalPower
Recommended values 8×5W radio 16 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-12: 8×5W RRH Power Recommended Values in TM8 (10 MHz bandwidth) Recommended values for a 8×5W RRH Power & 15MHz in TM8: Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port7port8PowerOffset cellDLTotalPower
Recommended values 8×5W radio 14 dBm 0.2 dB 0.2 dB -3.0 dB * -2.6 dB * -2.6 dB * -2.0 dB * -2.0 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-13: 8×5W RRH Power Recommended Values in TM8 (15 MHz bandwidth) Recommended values for a 8×5W RRH Power & 20MHz in TM8: Inputs
Recommended values 8×5W radio
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 45/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inputs referenceSignalPower primarySyncSignalPowerOffset secondarySyncSignalPowerOffset pBCHPowerOffset pDCCHPowerOffsetSymbol1 pDCCHPowerOffsetSymbol2and3 pCFICHPowerOffset pHICHPowerOffset paOffsetPdsch pbOffsetPdsch port7port8PowerOffset cellDLTotalPower
Recommended values 8×5W radio 13 dBm -0.8 dB -0.8 dB -3.0 dB * -3.3 dB * -3.3 dB * -2.0 dB * -1.5 dB * -3 dB 0 dB (PB = 1) 1.9 dB 37 dBm (5W)
Table 8-14: 8×5W RRH Power Recommended Values in TM8 (20 MHz bandwidth) Note that, since TLA6.0, 8×10W RRH is delivered and could be used for 2 carriers CA (each carrier with 8×5W) or 1 carrier only (8×10W) scenario, respectively. For the 2 carriers CA scenario, the recommended DL power setting is same as above 8×5W RRH power recommended values; For the 1 carrier only scenario, just plus 3dB on referenceSignalPower and cellDLTotalPower, respectively, based on above 8×5W RRH power recommended values. Based on field and analytical studies, ALU recommends: Changing paOffsetPdsch to dB-3 to improve Reference Signal SINR Adjusting CCCH’s and reference power offsets to maintain current amplifiers power utilization as well as to maintain PDSCH and CCCH’s SINR levels close to current level;
SCOPE: Improve Call Drop Rate.
8.2.6 CELL COVERAGE Examples on following settings: Central Frequency: 2.6 GHz RF channel models: Dense urban
8.2.6.1 UL CELL COVERAGE The Uplink cell coverage is defined as a target service that UE must satisfy at cell edge conditions. The worst case is 14.5 kbps which was selected for this case. The cell coverage is independent of channel bandwidth and is strictly dependent of UE Tx power which is 23dBm maximum for the UE’s supplied by third parties and tested by ALU. For TDD, cell coverage will also be impacted by UL/DL config. Hereunder is the test result for TDD config1. UL 2.6GHz Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 46/290
LTE Optimization Handbook TLA6.0 Target Service at cell edge Table 8-15: UL 2.6GHz
MGR/TIPS/NEA
PS14.5
Dense Urban in car Path Loss (dB) 115 Cell Range (Km) 0.70 Table 8-16: Path Loss & UL Cell Range in Dense Urban in Car (Field Results CMCC LST)
8.2.6.2 DL CELL OUTDOOR COVERAGE In Downlink the eNB’s ReferenceSignalPpower is influencing the cell range. Only 10dBm was selected for ReferenceSignalPower in the test. The coverage computation is independent of channel bandwidth. Environment ReferenceSignalPower (dBm) Cell Range (Km) Dense Urban in car 10 0.74 Table 8-17: DL Cell Range in Dense Urban in Car (Field Results CMCC LST) Almost every time the DL cell coverage will be higher than UL cell coverage. Not always selecting a high ReferenceSignalPower will mean better coverage: Hearing the Sync signals and MIB isn’t enough to obtain an RRC connected state.
8.2.6.3 TOTAL CELL COVERAGE The so called total cell coverage is expressed as a minimum between UL and DL. The thorough cell range is expressed by the following formula: Total cell coverage [m] = Min (UL cell range, DL cell range) The total cell range will almost every time be the UL one. In a dense site area where the cells are close to one another, and effects of shadowing, multipath propagation are very accentuated, cell coverage is limited and a high UE Tx Power will create interference in neighbouring cells when UE is inside it’s serving cell edge. The phenomenon is very common and needs to be analyzed carefully from the network planning stages. The Fractional Power control algorithm is a good way to improve conditions at cell edge by lowering the SIR target level so the interference in neighbouring cell is kept at a minimum.
8.2.6.4 PUSCH FRACTIONAL POWER CONTROL Fractional Power control is used in order to limit the interference that cell edge-users create to the neighbouring cells. In fractional power control, the transmit power adjustment pUSCHPowerControlAlphaFactor × PL compensates for only a fraction of the estimated path loss PL. The result is that the SINR achieved by the UE at the eNB varies linearly with the path loss. Higher levels of path loss are associated with lower SINR and vice versa. When the UE is close to the cell centre, the path loss decreases and hence the target SINR is increased. When the UE is at the cell edge, the path loss increases and hence the target SINR is decreased.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 47/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 8.2-3: Throughput for single UE vs. Path loss (Lab environment) In high path loss conditions, the throughput with a lower SIR target becomes better because eNB will grant a lower MCS but with more PRBs than for high SIR Target. For the Fractional power control tests, was used the following values /configuration expressed in the two examples given below. In the first example given the SIR target = 0 between Path loss 110 to 140dB. In the 2nd below the SIR target = 0 between Path loss 135 to 140dB. For pUSCHPowerControlAlphaFactor = 0.7 maxSIRtargetForFractionalPowerCtrl = 15.0dB minSIRtargetForFractionalPowerCtrl = 0.0dB uplinkSIRtargetValueForDynamicPUSCHscheduling = 15.0dB pathLossNominal = 60dB p0NominalPUSCH = -79 dBm Path loss where SIRtarget reaches 0dB: 110dB
For pUSCHPowerControlAlphaFactor = 0.8 maxSIRtargetForFractionalPowerCtrl = 15.0dB minSIRtargetForFractionalPowerCtrl = 0.0dB uplinkSIRtargetValueForDynamicPUSCHscheduling = 11.0dB pathLossNominal = 80dB p0NominalPUSCH = -85 dBm Path loss where SIRtarget reaches 0dB: 135dB
8.2.7 PUSCHPOWERCONTROLALPHAFACTOR Part of PUSCH power control and is intended to allow partial compensation of the path loss or otherwise stated it allows controlling, by decreasing, the PUSCH power for users in cell edge conditions. If set to 1; an increase in path loss will determine the same increase in PUSCH power. If this parameter is not set to 1, the increase in PUSCH power can be lower than the increase in path loss. It is thus a means of controlling the UL interference created in the neighbour cell by the UEs found near the cell edge. Because the value of this parameter represents a trade-off between minimizing interference and maximizing throughput, its value must be set according to the client’s desired network behaviour. If Fractional Power Control is used, the recommended value of this parameter should be [0.0, 1.0]. If Fractional Power Control is not to be used, the parameter must have the value 1. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 48/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Take care that this recommendation is only for commercial usage. Recommended & default Value = "1.0" when FPC is disabled; "0.8" when FPC is enabled.
NEA Recommended Value= “1.0” to reach peak UL performance in trial mode; but “0.0
Increasing the value of this parameter would: Increase PUSCH power for a given path loss. Increase the throughput for the all users. Increase the interference toward the neighbouring cells which might lower the throughput of users being in cell edge propagation conditions in neighbour cells. Decreasing the value of this parameter would: Decrease PUSCH power for a given path loss. Decrease the throughput for the all users. Decrease the interference toward the neighbouring cells which might increase the throughput of users being in cell edge propagation conditions in neighbour cells. KPI Impact: Coverage – higher values improve coverage Throughput – higher values will improve throughput, while lower values will decrease it. Capacity - higher values might reduce capacity, while lower values might increase it. The optimization process of this parameter should include the customer definition of the optimum trade-off between cell throughput and interference towards the neighbour cells. The choice can be different if cell wise optimization is to be performed or if a network wide setting is being aimed. Step 1: Set the pUSCHPowerControlAlphaFactor to 1 and connect the UE in Near-Cell radio conditions. For a closer view to the lab results IoT should be considered. Step 2: Start a UL UDP transfer and start driving from Near-Cell towards Edge-Cell. For a more consistent data we recommend a drive back as well logged in another trace. Step 3: Using the same cell and same route choose another value for pUSCHPowerControlAlphaFactor = {0.9, 0.8, 0.7, and 0.6} and repeat Step 1 and Step 2. Step 4: Post process the logged data and provide results in terms of UE TxPower, UL Throughput and PUSCH BLER.
8.2.8 QRXLEVMIN Clarifications regarding qRxLevMin: A parameter with this name appear in several objects and is then transmitted to UE inside several system information block types i.e.SIBs: CellSelectionReselectionConf – transmitted in SIB1 and SIB3 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 49/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
CellReselectionConfUtraFdd – transmitted in SIB6 CellReselectionConfUtraTdd – transmitted in SIB6 Note: For the moment, we only support mobility between the same frame structure, i.e. either FDD to FDD or TDD to TDD. So, which SIB6 will be broadcasted is based on currently used LTE frame structure. CellReselectionConfGERAN – transmitted in SIB7 The one that is object of this paragraph is transmitted in SIB1 which contains information relevant when evaluating if a UE is allowed to access a cell and defines the scheduling of other system information. This parameter impacts the cell size in terms of re-selection area i.e. mobility in idle mode. It configures the serving cell min required RSRP level used by the UE in cell reselection. The value sent over the RRC interface is half the value configured. Changing the value of this parameter will have an effect on the cell the UE is camped on during its idle mode. One way of optimizing it is to find the value that best superposes the cell size in idle with the cell size in active mode such that an idle-to-active transition would not result in an immediate handover decision. The exact selection criterion, S relev , is based on several values related to measured signal and power compensation level as below: Srxlev = Qrelevmeas - (Qrxlev min + Qrxlev min offser) - Pcompensation The selection is decided if
S rxlev 0
Pcompensation 0 and Qrxlev min offser is only considered when a periodic search for a higher priority PLMN is being performed. Thus, for a normal selection, the selection criterion is fulfilled if: Qrelevmeas - Qrxlev min > 0 or Qrelevmeas > Qrxlev min As long as the above relation is being satisfied the measured cell is selected. The variation of the cell size when various RSRP targets are set is given in the figure below. The information in this picture is only informative since the cell size variation strongly depends on the clutter. Recommended & Default Value= “-120” for 20MHz BW, TBD for other BWs. Increasing the value of this parameter would: Will lead the mobile to start cell-selection/re-selection procedure sooner because the inequality will be satisfied for a narrower range of measured values and then will artificially decrease cell size in idle mode. Indeed, for avoiding too many measurements to be performed for too long a time, there is a decision for starting inter-cell measurements based only on the received field level. Decreasing the value of this parameter would: Probably lead to unstable cell-reselection Allow UE to camp on the cell while being further away from the transmitting antennas.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 50/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Attach/Detach - low values might negatively impact (delay or make impossible) the attach operation. Coverage - lower value means larger cells. Mobility - high values might create coverage discontinuity in idle, as seen by mobile. When trying to match the idle mode cell size and active mode cell size, i.e. optimize the value of this parameter, drive tests must be performed in the cell. The testing procedure should comprise the following steps: Step 1: With UE in active mode, perform a drive test back and forth between the two cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 2: Post process the logged data and determine the cell edge, as being the positions at which the UE switched to the neighbour cell and the measured SINR at those locations. Step 3: Set the value of qRxLevMin to one of the following values {- 124, -122, -120, -118, -116, -114, -112, -110}. Step 4: With UE in idle mode, perform a drive test back and forth between the two cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 5: Choose another qRxLevMin and repeat Step 4. Step 6: Post process the logged data and determine the positions at which the UE started searching for another cell and the positions at which UE switched to the neighbour cell along with the measured SINR. Step 7: Based on the cell size in active mode and cell sizes in idle mode, choose the optimized value in order to compensate if you have a smaller or lager cell than you wish.
Distance (m)
Distance Vs qRxlevmin 500 450 400 350 300 250 200 150 100 50 0
Qrx_-110 Qrx_-112 Qrx_-114 Qrx_-116 Qrx_-118 Qrx_-120 Qrx_-122 Qrx_-124 Test route
Figure 8.2-4: Cell coverage (idle & connected) vs. qRxlevmin – 2600MHz (Field Results CMCC LST-Urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 51/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
UE in idle Vs UE in active Average Cell edge (m)
450 440 430 420 410
Ue in idle
400
Ue in active
390
380 370
Figure 8.2-5: DL coverage for UE in idle Vs UE in active (Field Results CMCC LST-Urban) According to the field test result above, we found that in Urban environment, the optimized value for qRxlevmin can be -118dBm. qRxLevMin selection if: Edge Cluster Cell Edge Cell (IM / MOB) Inter-Cells Gaps Inter-Cells
Overlapping Cell load
Figure 8.2-6: qRxLevMin Selection We can set different qRxLevMin for different cells respectively according to different scenarios and RF condition (e.g. tuning cluster cell edge, idle and active mobility cell edge matching, inter cells gap, inter cells overlap, cells load balance, etc.), just as show in Figure 8.2-6. The tuning purpose of qRxLevMin is to set suitable cell size, includes decreasing the gap between cells, controlling overlap of cells, matching the idle mode cell size and active mode cell size, etc..
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 52/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
8.2.9 P0NOMINALPUSCH This parameter is, somehow indirectly, impacting the power the UE transmits, before any power control commands is being received from the eNodeB. This parameter is a key RF optimization parameter. Higher settings will improve PUSCH reception, but will also drive higher UE Tx power leading to interference to neighbouring cells, and vice-versa. Its current default value is -96dBm. Indeed, the power of the UE will be adapted once the transmission is being started and the impact this parameter has on the UE power decreases with time. Pusch Power/RB = P0NominalPusch + alpha*Pathloss Optimization would mean finding the best value that, at the same time, for which the PUSCH reception is good enough, even at the beginning of the PUSCH transmission, and the interference created towards the neighbour cells is kept to an acceptable level. For more details on the PUSCH power control see the end of this paragraph. Recommended & Default Value= “-96” in case of fractional power control not used; Default Value=“-82” in case of fractional power control with 0.8 value in pUSCHPowerControlAlphaFactor. Expected behaviour when value is modified Increasing the value of this parameter would: Increase the interference in the neighbour cells at the beginning of PUSCH transmission which might temporally decrease the throughput of users found at the cell edge in the neighbour cells. Temporally result in a high SINR for PUSCH transmission which can be reflected in higher MCSs at the beginning of PUSCH transmission. The convergence of the PC algorithm could take more time thus the interference towards other cells could last longer Decreasing the value of this parameter would: Temporally result in a low SINR for PUSCH transmission which can be reflected in higher BLER and/or lower MCSs at the beginning of PUSCH transmission.
KPI Impact: Coverage – higher value will temporarily increase the UL interference Access – lower values might increase the access time
For optimizing the value of this parameter for minimizing the interference in neighbour cells, two cells and several UEs are needed (in the interfering cell). The following steps must be performed (this recommendation is only for the UL Fractional Power Control Disabled): Step 1: In victim sell, use the default p0NominalPUSCH value. In the interfering cell, set p0NominalPUSCH to one of the values {-108, -104, -100, -96, -92, -88, -84, -80}. Step 3: In interfered cell perform an UL full buffer data transfer and log data related to PC command (TPC command field or F value) and PUSCH BLER and PUSCH (DM RS) SINR. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 53/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 4: In the interfering cell, with UEs located near the cell edge, perform UL full buffer data transfer with several UEs, in synchronous manner. Step 5: In the interfering cell choose another value for p0NominalPUSCH and repeat Step 3 and Step 4. Step 6: Post process the logged data and provide results in terms PUSCH BLER of the average PUSCH DM RS SINR for each value p0NominalPUSCH in the interfering cell.
FFS: The field test for p0NominalPUSCH tuning still has not been performed yet, the results will be provided after the test cases are implemented. The most important power control aspects are: Open-loop power control with slow aperiodic closed loop correction factor Fractional path loss compensation with PL compensation factor Accumulated UE-specific closed-loop correction is used. The setting of the UE Transmit power PPUSCH for the physical uplink shared channel (PUSCH) transmission in the subframe is defined by:
PPUSCH (i) = min { PMAX, 10.log10(MPUSCH(i)) + P0_PUSCH + pUSCHPowerControlAlphaFactor.PL + ΔTF(TF(i)) + f(i) } where PMAX is the maximum allowed power that depends on the UE power class MPUSCH(i) is the bandwidth of the PUSCH transmission expressed in number of resource blocks taken from the resource allocation valid for uplink subframe i from scheduling grant received on subframe i-KPUSCH. P0_PUSCH is a parameter obtained as a sum of a cell specific nominal component p0NominalPUSCH signalled from higher layers and a UE specific component p0UePUSCH. pUSCHPowerControlAlphaFactor is a cell specific parameter signalled from higher layers in order to support fractional power control. PL is the downlink path loss estimate calculated in the UE. ΔTF(TF(i)) denotes the power offset depending on PUSCH transport format TF(i). Both accumulated and non accumulated power control rules are used – this is set by means of parameter accumulation Enabled. The current PUSCH power control adjustment state in subframe i is given by f(i): f(i) = f(i-1) + δPUSCH(i-KPUSCH), if accumulation Enabled is enabled f(i) = δPUSCH(i-KPUSCH), if accumulation Enabled is disabled where δPUSCH is a UE specific correction value in dB, also referred to as a TPC command and is included in PDCCH with DCI format 0 on subframe i-KPUSCH. f(0) = 0. For case when enabling Fractional Power control use following formula for applying the correct value to the parameter: P0NOMINALPUSCH = SIRTARGETMIN + (1 - Alpha) x PLMAX + IIOT Example 1: for pUSCHPowerControlAlphaFactor =1, and SINR_target_nominal = 1 dB, p0NominalPUSCH = 1 + 0 -112 = -111 dBm Example 2: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 54/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
for pUSCHPowerControlAlphaFactor =0.7, and SINR_target_nominal = 15 dB, p0NominalPUSCH = 15 + (1-0.7)*140 – 112 = -55 dBm Where SINR_target_nominal = uplinkSIRtargetValueForDynamicPUSCHscheduling
Figure 8.2-7: Po_pusch_Nominal Impact
8.2.10 UPLINKSIRTARGETVALUEFORDYNAMICPUSCHSCHEDULING This parameter is used inside PUSCH power control algorithm as outer loop power control for nonsemi-static, which means the PUSCH Tx power is adapted through the closed-loop power control and tries to keep the actual UL SINR close to the target UL SINR. It is used as an initial target for the SINR values. During transmission, the SINR targets are changing based on the measured path loss. The input of the UL outer-loop power control function is the path loss along with some other parameters. The SIR target is modified by using the following formula:
SIRNew_Target _PUSCH SIRTarget_PUSCH_initial 1 pUSCHPowerControlAlphaFactor ( PLav pathLossNominal) max min minSIRtargetForFractionalPowerCtrl maxSIRtargetForFractionalPowerCtrl The value of uplinkSIRtargetValueForDynamicPUSCHscheduling will also be the first value of SIRTarget_PUSCH_initial in the iteration above. As with the previous parameter, the value of SIR target is bounded. The value depends on a Power control factor and a nominal pathloss, both being parameters that can be set in the database. Note that, IRC is kept in TLA6.0 as well as in TLA4.0. When using IRC, besides the original SINR, one new SINRforIRC is reported from L1, ULS should choose to use one of them based on the flag l1ReceiverMethod from MIM. For MRC, the original SINR is still used.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 55/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
In the trial mode, the higher the setting of uplinkSIRtargetValueForDynamicPUSCHscheduling the higher the UL throughput, but also the higher the interference generated in the neighbouring cells. However, as in this mode, the focus is on the target cell performance rather than the overall network performance, this is not an issue. For trial mode, the nominal SINR target uplinkSIRtargetValueForDynamicPUSCHscheduling is defaulted to 16.0 in case of fractional power control not used (i.e. fractional power control used with 1.0 value in pUSCHPowerControlAlphaFactor). In the commercial mode, and on a network level, the higher the SINR target (i.e. the higher the setting of uplinkSIRtargetValueForDynamicPUSCHscheduling) the higher the near-cell throughput but the higher the interference generated in the different cells of the network (and thus the lower the cell-edge throughput and at some point the lower overall cell throughput too). In this case, the default setting of this parameter should be as follows:
Recommended Value= “16.0” in case of fractional power control not used (i.e. fractional power control used with 1.0 value in pUSCHPowerControlAlphaFactor); Or Recommended Value = “14.0” in case of fractional power control used with 0.8 value in pUSCHPowerControlAlphaFactor Or Recommended Value = “14.0” in case of fractional power control used with 0.7 value in pUSCHPowerControlAlphaFactor Actually, this parameter is tightly linked with AlphaFactor and the pathLossNominal, fine-tuning is required to achieve the right level of interference. Ideally, the tuning would be done on a cell-bycell basis (as the topology and the resulting radio propagation environment generally change from cell to cell). Increasing the value of this parameter would: Increase the transmission power of the UE which would result in using higher MCSs and obtaining higher throughputs. If the default setting of pUSCHPowerControlAlphaFactor is used, then the SIR target will be the same irrespective of the path loss and the UE power will be kept high. The UE power will reach its maximum value for lower path losses and thus the life of UE battery will be decreased. Reach the highest UE power for a lower pathloss which, in extremis, would limit the coverage. Increase interference in the neighbour cells due to higher transmitting power. This would result in lower throughputs in the neighbour cells. Decreasing the value of this parameter would: Decrease the power of the UE which would result in lower MCSs and lower throughputs. Decreasing this parameter will decrease the overall level of interference and hence improve the throughput of cell-edge users at the expense of cell-centre UEs, i.e. the peak throughputs will be lower.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 56/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Throughput: high values increase the throughput for near cell and mid-cell conditions. Capacity - high values allow reaching the capacity for a wider range of propagation conditions. Coverage - high values might reduce the coverage if the target is not dynamically adjusted based on propagation conditions. Mobility - might negatively impact the throughput during handover if the threshold is set too high. For optimizing the value of this parameter for minimizing the interference in neighbour cells while maximizing the throughput in the analyzed cell, two cells and several UEs are needed (in the interfering cell). The following steps must be performed: Step 1: In victim cell, use the default parameters. In the interfering cell, set uplinkSIRtargetValueForDynamicPUSCHscheduling to one of the values {16, 14, 12, 10, 8}. Step 2: In victim cell perform an UL data transfer and log data related to PC command (TPC command field or F value) and PUSCH BLER and PUSCH DM RS SINR. Step 3: In the interfering cell, with UEs located near the cell edge, perform UL full buffer data transfer with several UEs, in synchronous manner and log the value of the throughput. Step 4: In the interfering cell choose another value for uplinkSIRtargetValueForDynamicPUSCHscheduling and repeat Step 2 and Step 3. Step 5: Post process the data and choose the value of uplinkSIRtargetValueForDynamicPUSCHscheduling that provides an acceptable trade-off between the throughput in the interfering cell and the throughput in the victim cell.
-( PU 1n Co er ow
HP SC
uplinkSIRtargetValueForDynamicPUSCHscheduling
=
maxSIRtargetForFractionalPowerCtrl
pe s lo
Target SINR
lp
lA
tro ha c Fa r)
to
minSIRtargetForFractionalPowerCtrl
pathLossNominal
PL
Figure 8.2-8: Slope - PuschPowerControl vs. uplinkSIRtargetValueForDynamicPUSCHscheduling
8.3 FEATURE LINKED
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 57/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
8.3.1 T115590- SUPPORT OF MULTI-RRH PER CELL (ONE LOGIC CELL) FOR INDOOR COVERAGE 8.3.1.1 HIGH LEVEL DESCRIPTION AND BENEFITS: Traffic and Capacity in Hotspots are increased very fast, which makes it very important to improve coverage and capability in Hotspots, this is real challenge. Indoor scenario is a typical Hotspot with high traffic and capacity requirements but hard to provide good coverage and satisfied QoE. Dense Outdoor Hot Spots
Residential
Indoor Hot Spots
Rural Areas
Figure 8.3-1: Typical wireless network example
Figure 8.3-2: Key challenges of wireless network - coverage and capacity The feature T115590 Multi-RRH in one cell can be applied for indoor environment typically. It can enlarge the coverage of indoor within one logical cell and reduce frequent handover opportunity. And this feature also can be applied for the tiding service scenario between different areas. This is also for CMCC requirements on multi-RRH connected to one cell feature for indoor deployment. In order to satisfy customer’s requirements, we plan to support multi-RRH with one cell in TLA6.0. One BBU with 1~3 bCEM boards, up to 4 RRH 2x20w (10M or 20M) for one cell (1 bCEM); up to 6 RRH2x20w (10M or 20M) for one BBU.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 58/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 8.3-3: One logical cell typical architecture
8.3.1.2 HOW TO ACTIVATE: The feature is license controlled by O&M with parameter isMultiRrhEnabled set to “true”. For feature activation, the initial step is to ensure that the following parameters are set as exemplified: Object CellActivationService
LteCellTDD
Attribute isMultiRrhEnabled isInterTransmissionModeSwitchingEnable d spare3 . bit #2 (isOneLogicCellEnabled)
CellL1ULConfTDD
antSubArrayGroupingScheme
CellActivationService
CellL1ULConfTDD l1ReceiverMethod LteCellTDD numberOfDLAntennas LteCellTDD numberOfULAntennas LteCellTDD transmissionMode Table 8-18: Parameters to activate feature
Value true disabled 1 (true) {0,1,2,3,4,5,6,7 } MRC dlAntenna2 ulAntenna2 TM2/3/4
8.3.1.3 FEATURE IMPACTS ON ENB & TUNABLE PARAMETERS: Feature Impacts on eNB In TLA6.0, the architecture for this feature shall be based on 1S8A package on bCEM, DLS and ULS are performed as 2A case. The main eNB impacted Subsystems are summarized below: Callp: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 59/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Add RRH index information into Msg3/Msg4 L2 DL scheduler: Same as 2A processing + add 1 bitmap in scheduling info (RRH index for ACK/NACK receiving on PUCCH) UE RRH index table, based on PRACH or SRS report Use 2A power setting, ignore real antenna number Reset to initial status if change of RRH index (SINR, RI, CQI, MCS, BLER control…) L2 UL scheduler: UE RRH index table from L1 report, PRACH or SRS Power Control, ulTxPowerIncreasedFor2Rrh – add on target SINR or subtracted from estimated SINR used by power control if multi RRH RRH index change – reset to initial status (PUCCH SINR, PUSCH SINR, SRS SINR…) L1: RRH area detection modules are added in SRS processing and selected RRH index(es) for each UE shall be reported to L2 ULS Related thresholds about PUSCH processing shall be updated according to the number of receive antennas for the UE For PUCCH receiver in DSP, perform PUCCH signal as 2A or 4A case for each UE and corresponding thresholds, shall also be read from corresponding tables (2A table or 4A table) according to RRH index information of the UE. DSP also handles the processing of ACK/NACK, PCQI/SR on PUCCH based on RRH index information (got from SRS processing or PRACH reporting). The detail impacts on eNB please refer to doc [4]. This feature focuses on improving coverage and capacity for Indoor LTE deployments. Tuneable Parameters Ideally some exercise should be defined for tuning some of the parameters… in this case; (default values presented below). Object MultiRrh LteCellTDD
Attribute
Value Range
areaIsolationForMultiRrh ulTxPowerIncreasedFor2Rrh spare4 . bits #28~29 (servingRrhPowerGapMargin)
[0,…,9] [0.0,…,5.0]
Default Value 6 0.0
[1,2,3]
2
Table 8-19: Tuneable parameters
9 ACCESS OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is access in LTE. Normally some questions arise, such as: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 60/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
When to perform access optimization? What method to apply? Which parameters can help improving access to the network?
Mainly the Access optimization can occur when the attach success rate is below the ALU KPI As main indicator to evaluate the performance several tests to access the network should be performed; although before starting playing with the parameterization; usually is part of best practice rules for in Near Cell /Mid Cell & Cell Edge test to follow up simple steps as: Check the CQI Evaluate RSSI vs. SNR relation Evaluate RSRP vs. RSRQ
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If regardless of the correct values, still facing some performance issues, the below parameters can be used in order to correct the situation. When changing parameters; you can adopt a more error-free approach, meaning that a parameter is changed at each time. If three or four parameters are changed same time… it could be difficult to understand which one is bringing the improvement in performance. As note; please remember that this can be a static test in each position, or can be a moving test… the same principles can be applied in both situations.
9.1 PARAMETERS OPTIMIZATION FOR IMPROVING ATTACH/DETACH PROCEDURES The attach procedure is one of the most basic procedure in LTE. It implies an exchange of several messages between eNodeB and UE. The physical channel that is used for attach procedure is PRACH. The UE transmits a Random Access preamble when communication with the E-UTRAN is needed, for example in order to make a service request. When the eNodeB correctly receives a RA preamble, and if resources are available, it will respond with a RA response. This RA response contains, among other information, an initial UL grant to be used by the UE for the next UL message, namely the RRC Connection Request message (see Figure 12.1-12). There are several parameters directly involved in the RACH preamble procedure. The most important of them are described in this chapter.
9.1.1 PREAMBLEINITIALRECEIVEDTARGETPOWER Open-loop power control is applied for initial transmission of RACH (i.e. message1). The transmit power is determined by taking into account the total UL interference level and the required SINR operating point. Transmit power can be determined at the UE as:
(1) PRACH_msg1 min {PCMAX, PL P0 _PREAMB LE Δ PREAMB LE (N PREAMB LE 1) Δ RAMP_UP } Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 61/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
• The term PL is the DL path loss estimated at the UE from DL RS. • P0_PREAMBLE is the preamble received power set point determined at the eNodeB. This parameter is calculated from the target SINR operating point (SINRTarget), and the UL interference-plus-noise (IN) power in the PRACH resource. Possible margin may be added to account for any measurement error. The parameter preambleInitialReceivedTargetPower configures P0_PREAMBLE.
( 2) P0 _PREAMBLE SINRT arg et IN M arg in} • PREAMBLE is the power offset value dependent on PRACH preamble format which is given by prach-ConfigIndex. The preamble format based power offset values are presented in Table below: Preamble Format 0 1 2 3 4
DELTA_PREAMBLE value 0 dB 0 dB -3 dB -3 dB 8 dB
. • RAMP_UP is the power ramping step size. It is configured by parameter preambleTransmitPowerStepSize. • N PREAMBLE is the preamble transmission number (=1 for first transmission) up to maximum number of transmissions. This maximum number of transmissions is configured by parameter preambleTransMax. As show in Figure 9.1-1 the power of 1st preamble transmission need to fulfil the configured initial target received power, if eNB cannot decode the 1st received preamble, UE will retransfer preamble after a timer. At each new transmission of the preamble, the power is ramped up by Δ RAMP_UP dB and the trans number n is updated as n+1, until we reach the maximum number of transmissions preambleTransMax. If the mobile still does not receive any Random Access response from the eNodeB, the UE MAC layer then declares the Random Access procedure as failed.
Figure 9.1-1: preambleTransMax vs. preambleInitialRceivedTargetPower vs. preambleTransmitPowerStepSize
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 62/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Po_PREAMBLE inpact on UE TxPow vs PL(RA) (TRY1) 30
20
PRACH TxPower
10
0 Try1#-104 Try1#-96 -10
-20
-30
-40 70
75
80
85
90
95
100
105
110
115
120
125
130
135
140
145
150
-37.5
-42.5
-47.5
-52.5
-57.5
-62.5
-67.5
-72.5
-77.5
-82.5
-87.5
-92.5
-97.5
-102.5
-107.5
-112.5
-117.5
PL Estim. RSRP
Figure 9.1-2: Po_preamble impact on UE Tx Power vs. PL(RA) (TRY1) Based on the above results in Figure 9.1-2 we can conclude that the higher the pathloss the higher the UE Tx power for preamble transmission until the pathloss grows to a point (e.g. 120dB).
Figure 9.1-3: Po_preamble impact on UE Tx Power vs. PL (RA) Based on the above results in Figure 9.1-3 we can also conclude that the higher the preambleInitialReceivedTargetPower the higher the UE Tx power for preamble transmission in the same pathloss position, the Tx power also increases according to the preambleTransmitPowerStepSize between try 1#, try 2# and try3# until the pathloss grows to a point which needs UE transmit preamble with Pmax.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 63/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 9.1-4: preambleTransMax vs. preambleInitialRceivedTargetPower vs. preambleTransmitPowerStepSize (example of values) Let’s consider PL = 90;
PRACH_msg1 min{PCMAX , PL P0_PREAMBLE PREAMBLE ( N PREAMBLE 1) RAMP_UP}
; 1st try
N PREAMBLE =1 PRACH_msg1= min {PMAX, 90 + P0_PREAMBLE + 0} => PRACH_msg1= -14 [dBm] PRACH_msg1= min {PMAX, 90 + P0_PREAMBLE + 6} => PRACH_msg1= -8 [dBm] PRACH_msg1= min {PMAX, 90 + P0_PREAMBLE + 12} => PRACH_msg1= -2 [dBm] Recommended & Default Value= “dBm-94” for trial mode.
Expected behaviour when changing this parameter Increasing the value of this parameter would: Minimize the repetitions i.e. RACH attempts and hence expedite call setup, but will cause higher interference to other cells during the attach procedure. Thus there will be higher interference during a shorter period of time. Decreasing the value of this parameter would: Increase RACH repetitions / call setup delay and decrease the interference. One specific UE will generate lower interference but for a longer period of time. When there are several UEs in the cell, the total interference might not decrease.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 64/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Attach/Detach: low values of this parameter can delay the success of the attach operation. Very low values might even make the attach operation impossible. Mobility: low values might lengthen the interruption time.
A good optimization criterion for this parameter would be to set it to the lowest limit that ensures that the required RACH preamble success rate at 1st attempt is achieved. The optimization of this parameter should consider the steps below: Step 1: In the database, check that the default value dBm-94 is set. Step 2: Perform drive tests while connecting and disconnecting the UE in near-cell, mid-cell and cell-edge conditions of the cell while logging the data and the GPS position. Step 3: Change the value of the parameter to the following {dBm-104, dBm-102, dBm-100, dBm-98, dBm-96} and repeat Step 2 on the similar route. Step 4: Provide results in form of three graphs representing: •Average number of repetitions before RA success vs. preambleInitialReceivedTargetPower •Average preamble power for the successful try vs. preambleInitialReceivedTargetPower. •Average time for RA success vs. preambleInitialReceivedTargetPower
9.1.2 PREAMBLETRANSMITPOWERSTEPSIZE This parameter is a key RF optimization parameter that impacts connection setup performance and uplink interference to neighbouring cells. Higher values will minimize the repetitions/ RACH attempts and hence expedite connection setup, but will cause higher interference to other cells. Lower values will tend to increase RACH repetition/ connection setup delay. The current default value for this parameter is dB6.
Recommended & Default Value= “dB6” Increasing the value of this parameter would: Minimize the repetitions i.e. RACH attempts and hence expedite call setup, minimize the time the UE generates interference in the system due to random access procedure. Highest values might create unnecessary-high interference for the last random access attempt (the successful one). Decreasing the value of this parameter would: Increase RACH repetitions / call setup delay and decrease the interference. One specific UE will generate lower interference but for la longer period of time.
KPI Impact: Attach/Detach: low values of this parameter can delay the success of the attach operation. Very low values might even make the attach operation impossible. Mobility: low values might lengthen the interruption time. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 65/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
A good optimization criterion for this parameter would be to set it to the value that minimizes the number of preamble transmissions. Indeed, the size of the power ramp-up step is a parameter that can decrease the number of failed attachment attempts but would not increase the success of the first attempt. The optimization of this parameter must be performed in conjunction with the optimization of the previous parameter. It is best to start optimization by first considering the highest value of this parameter i.e. dB6 The optimization of this parameter should consider the steps below: Step 1: In the database, check that the value dB6 is set. Step 2: Perform drive tests in near-cell, mid-cell and cell-edge conditions while connecting and disconnecting the UE and while logging the data and the GPS position. Step 3: Change the value of the parameter to the following {dB4, dB2} and repeat Step 2 on the similar route. Step 4: Provide results in form of three graphs representing: •Average number of repetitions before RA success vs. preambleTransmitPowerStepSize •Average preamble power for the successful try vs. preambleTransmitPowerStepSize. •Average time for RA success vs. preambleTransmitPowerStepSize
NC
20.0 15.0 10.0 5.0 0.0 -5.0 -10.0 -15.0 -20.0
MC CE
Linear (NC) Linear (MC)
(-104,6)
(-104,4)
(-104,2)
(-102,6)
(-102,4)
(-102,2)
(-100,6)
(-100,4)
(-100,2)
(-98,6)
(-98,4)
(-98,2)
(-96,6)
(-96,4)
(-96,2)
(-94,6)
(-94,4)
(-94,2)
Prach Txpower(dbm)
(preambleInitialReceivedTargetPower,preambleTrans mitPowerStepSize) VS Prach Txpower
Linear (CE)
preambleInitialReceivedTargetPower,preambleTransmitPowerStepSize Figure 9.1-5: preamble TxPower vs. preambleInitialRceivedTargetPower Vs preamblePowerStepSize (Field Results CMCC LST-Urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 66/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
(preambleInitialReceivedTargetPower,preambleTrans mitPowerStepSize) VS RA repetition NC RA repetition
2
MC CE
1
Linear (NC) Linear (MC) Linear (CE)
0 (-104,6)
(-104,4)
(-104,2)
(-102,6)
(-102,4)
(-102,2)
(-100,6)
(-100,4)
(-100,2)
(-98,6)
(-98,4)
(-98,2)
(-96,6)
(-96,4)
(-96,2)
(-94,6)
(-94,4)
(-94,2)
preambleInitialReceivedTargetPower,preambleTransmitPowerStepSize Figure 9.1-6: preamble Re-transfer Number vs. preambleInitialRceivedTargetPower Vs preamblePowerStepSize (Field Results CMCC LST-Urban)
RA Latency (ms)
(preambleInitialReceivedTargetPower,preambleTrans mitPowerStepSize) VS RA Latency NC
80 75 70 65 60 55 50 45 40
MC CE Linear (NC) (-104,6)
(-104,4)
(-104,2)
(-102,6)
(-102,4)
(-102,2)
(-100,6)
(-100,4)
(-100,2)
(-98,6)
(-98,4)
(-98,2)
(-96,6)
(-96,4)
(-96,2)
(-94,6)
(-94,4)
(-94,2)
Linear (MC) Linear (CE)
preambleInitialReceivedTargetPower,preambleTransmitPowerStepSize Figure 9.1-7: Near Cell RA Success Rate vs. preambleInitialRceivedTargetPower (Field Results CMCC LST-Urban)
9.1.3 SCHEDULED TRANSMISSION (DELTAPREAMBLEMSG3 OR TPCRACHMSG3) The Nominal transmit power for RACH msg3 which carry RRC connection request message while RRC connection establish, denoted as PO_NOMINAL_PUSCH (2) is computed at the UE as:
P
( 2) P
O_NOMI- NAL_ PUSCH - Solely for authorized O_PREAMBpersons LE PREAMB LE_M sg3 Proprietary - Use pursuant to Company instruction Alcatel-Lucent Confidential having a need to know
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 67/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
where PREAMBLE_Msg3 is the nominal power offset between RACH preamble and RACH message 3. It is configured by parameter deltaPreambleMsg3 as follows
PREAMB LE_Msg3 deltaPream bleMsg3 3dB
. The Transmit power of the UE for RACH message 3 is determined (in dBm) by normal PUSCH power control formula:
PPUSCH (i ) min{ Pmax ,10 log10 ( MPUSCH (i )) P0 _ NOM INAL_ PUSCH ( 2) P0 _ UE _ PUSCH ( 2) pUSCHPowerControlAlphaFactor PL TF (i ) f (i )}
f (i) is initialized (for the first transmission of RACH message 3) as follows f (0) Prampup msg2 where
msg2 is the TPC command indicated in the Random Access Response (RACH message 2) and
set by tPCRACHMsg3 (see next parameter for optimization) with the TPC command and parameter value mapping table below: tPCRACHMsg3 -6 -4 -2 0 2 4 6 8
TPC Command 0 1 2 3 4 5 6 7
Prampup corresponds to the total power ramp-up from the first to the last preamble. For subsequent transmission of RACH message 3, accumulated power control formula f (i) f (i 1) PUSCH (i K PUSCH ) applies. deltaPreambleMsg3 – mostly impacts the first transmission of Msg3, before TPC acts for adjusting the transmitting power according to propagation conditions.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 68/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 9.1-8: Parameters dependency and relations For TDD UL/DL configurations 1-6, K PUSCH is given in 1 For TDD UL/DL configuration 0: If the PUSCH transmission in subframe 2 or 7 is scheduled with a PDCCH of DCI format 0 in which the LSB of the UL index is set to 1, K PUSCH 7 ; For all other PUSCH transmissions, K PUSCH is given in 1. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 69/290
LTE Optimization Handbook TLA6.0
0
0 -
1 -
subframe number i 2 3 4 5 6 7 6 7 4 - - 6
1
-
-
6
4
-
-
-
6
4
-
2
-
-
4
-
-
-
-
4
-
-
3
-
-
4
4
4
-
-
-
-
-
4
-
-
4
4
-
-
-
-
-
-
5
-
-
4
-
-
-
-
-
-
-
6 - for TDD configuration 0-6
7
7
5
-
-
7
7
-
TDD UL/DL Configuration
Table 9-1 K PUSCH
MGR/TIPS/NEA
8 7
9 4
Note that, since TLA2.1, the Power Control of first transmission of MSg3 depends on the power measurement on the preamble of msg1. Retransmissions of msg3 (due to HARQ) will not get an additional Power control. Which means adaptive retransmission (containing TPC) for Msg3 is not supported. Increasing the value of these parameters would: Minimize the number of HARQ retransmissions and hence shorten attach procedure time Minimize the time the UE generates interference in the system due to random access procedure. Interference is less critical for Message3 due to the fact that it is scheduled. Decreasing the value of these parameters would: Potentially increase the number of HARQ retransmissions with lower initial power. The UE will generate lower interference but possibly for a longer period of time. KPI Impact: Attach/Detach: low values of this parameter(s) can slightly delay the success of the attach operation.
9.1.4 DELTAPREAMBLEMSG3 A good optimization criterion for this parameter (deltapreamblemsg3) would be to set it to the value that minimizes the number of HARQ retransmissions of Msg3.
Recommended & Default Value= “12” The optimization of this parameter should consider the steps below: Step 1: In the database, check that the value 12 is set. Step 2: Perform drive tests in near-cell, mid-cell and cell-edge conditions while connecting and disconnecting the UE and while logging the data and the GPS position. Step 3: Change the value of the parameter to the following {2, 4} and repeat Step 2 on the similar route. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 70/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 4: Provide results in form of one graph representing: • Average number of HARQ retransmissions vs. deltaPreambleMsg3 Step 5: Consider as optimum the minimum value of deltaPreambleMsg3 which requires the minimum number of HARQ retransmission of Msg3.
9.1.5 TPCRACHMSG3 A good optimization criterion for this parameter (tPCRACHMsg3) would be to set it to the value that minimizes the number of HARQ retransmissions of Msg3.
Recommended & Default Value= “4dB” The optimization of this parameter should consider the steps below: Step 1: In the database, check that the value 4dB is set. Step 2: Perform drive tests in near-cell, mid-cell and cell-edge conditions while connecting and disconnecting the UE and while logging the data and the GPS position. Step 3: Change the value of the parameter to the following {2dB, 6dB} and repeat Step 2 on the similar route. Step 4: Provide results in form of one graph representing: • Average number of HARQ retransmissions vs. tPCRACHMsg3 Step 5: Consider as optimum the minimum value of tPCRACHMsg which requires the minimum number of HARQ retransmission of Msg3.
Random Access Success Rate (%)
(deltaPreambleMsg3 ,tPCRACHMsg3) Vs Ra Success Rate 100 80 60 nc
40
mc
20
ce
0 (0,0) (0,2) (0,4) (0,6) (2,0) (2,2) (2,4) (2,6) (4,0) (4,2) (4,4) (4,6)
deltaPreambleMsg3 ,tPCRACHMsg3(db)
Figure 9.1-9: RA Success Rate Vs deltaPreambleMsg3 Vs tPCRACHMsg3 (Field Results CMCC LSTUrban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 71/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Puschpower(dbm)
(deltaPreambleMsg3 ,tPCRACHMsg3) VS Puschpower 30 20 10
nc
0 -10
mc (0,0) (0,2) (0,4) (0,6) (2,0) (2,2) (2,4) (2,6) (4,0) (4,2) (4,4) (4,6)
-20
ce
deltaPreambleMsg3 ,tPCRACHMsg3(db)
Figure 9.1-10: PUSCH TxPower Vs deltaPreambleMsg3 Vs tPCRACHMsg3 (Field Results CMCC LSTUrban)
10 DOWNLINK THROUGHPUT OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is Downlink Throughput in LTE. Normally some questions arise, such as: When to perform DL Throughput optimization? What method to apply? Which parameters can help improving DL Throughput?
Mainly the DL t-put optimization can occur when the average throughput value for a specific location is not matching the ALU product specification for a determined Bandwidth target. Before starting playing with the parameterization; usually is part of best practice rules for in Near Cell /Mid Cell & Cell Edge test to follow up simple steps as: Check the CQI Check the MCS triggered Check the DL BLER Evaluate RSSI vs. SNR relation Evaluate RSRP vs. RSRQ
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If regardless of the correct values, still facing some performance issues, then below parameters can be used in order to correct the situation. When changing parameters; you can adopt a more error-free approach, meaning that a parameter is changed at each time. If three or four parameters are changed same time… it could be difficult to understand which one is bringing the improvement in performance. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 72/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
As note; please remember that this can be a static test in each position, or can be a moving test… the same principles can be applied in both situations. Note that, we suggest doing the throughput (or KPI) performance or parameters optimization after RF optimization of the whole network or at least some cluster had been well done.
10.1 PARAMETERS OPTIMIZATION FOR IMPROVING DOWNLINK THROUGHPUT 10.1.1 DLMCSTRANSITIONTABLE This table contains 28 float values representing the thresholds of SINRs values for which the DL modulation is being changed and it is part of an intricate algorithm inside DL scheduler. Optimization of this table would imply changing the values of the thresholds either by decreasing them or by increasing them. Indeed, it is not necessary to have them all increased or all decreased. Both uniform and non-uniform modifications of these values are possible. Below, several suggestions are presented, two of them implying uniform modifications of threshold values and four of them considering non-uniform modifications.
Figure 10.1-1: Radio link Quality vs. MCS Robustness vs. Throughput In the above Figure 10.1-1 it can be observed that both Radio Link Quality, MCS’s Robustness and Throughput are closely related… meaning that for a better Radio Link, this would imply a less robust MCS, but in other hand, the final result is a higher Throughput!
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 73/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-2: Radio link Quality vs. dlMCSTransition Table vs. Throughput In the table below there are several (academic) examples of threshold tuning for a 10MHz band. Default
Down Shift
Up Shift
-2.5 -2.00 -1.25 -0.50 0.50 0.75 2.00 2.75 3.75 4.75 5.00 6.00 7.00 7.25 8.25 9.00 10.25
-2.5 -2 -1.25 -0.25 0.75 1.75 2.5 3.5 4.5 5.5 6.25 6.75 7.75 8.5 9.5 10.5 11
-2.5 -4 -3.75 -3.5 -3.25 -2.25 -1.5 -0.5 0.5 1.5 2.25 2.75 3.75 4.5 5.5 6.5 7
DecreaseLower IncreaseHigher Significant -2.5 -4 -3.74 -3.67 -3.44 -3.22 -2.91 -2.61 -2.16 -1.71 -1.26 0 2.51 6.51 10.13 12.84 14.65
DecreaseLower IncreaseHigher Moderate -2.5 -4 -3.495 -2.96 -2.345 -1.735 -1.205 -0.555 0.17 0.895 1.495 2.375 4.13 6.505 8.815 10.67 11.935
IncreaseLower DecreaseHigher Significant -2.5 -3.89 -1 1 2.58 4 4.88 5.45 5.83 6.13 6.21 6.28 6.36 6.5 6.88 6.96 7.11
IncreaseLower DecreaseHigher Moderate -2.5 -3.945 -2.125 -0.625 0.665 1.875 2.69 3.475 4.165 4.815 5.23 5.515 6.055 6.5 7.19 7.73 8.055
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 74/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
DecreaseLower DecreaseLower IncreaseLower IncreaseLower IncreaseHigher IncreaseHigher DecreaseHigher DecreaseHigher Significant Moderate Significant Moderate 11.50 12 8 15.63 12.925 7.26 8.63 11.75 12.5 8.5 16.23 13.555 7.49 8.995 12.50 13.5 9.5 16.53 14.25 7.8 9.65 13.25 14.25 10.25 16.83 14.73 8.18 10.215 14. 25 15.25 11.25 17.06 15.305 8.57 10.91 15.25 16 12 17.29 15.755 9.26 11.63 16. 25 17 13 17.36 16.33 9.72 12.36 16.75 17.25 14 17.44 16.905 10.41 13.205 17.75 17.5 14.75 17.51 17.14 11.41 14.08 18.75 17.75 15.75 17.59 17.36 13.32 15.535 0.00 18 16 17.66 18 17.54 17.77 Table 10-1: Examples of threshold tuning for a 10MHz band (academic only, not applied in any trial /project). Default
Down Shift
Up Shift
Recommended & Default Value= "Default Table" Note: The dlMCSTransitionTable for 10MHz and dlMCSTransitionTable for 20MHz bandwidth are different. Increase all threshold values (up-shift) will result in lower data rates because higher MCS will only be selected for higher values of SINRs. Indeed, due to improved SINRs when a given MCS is selected, there will be a lower percentage of transmission errors over air interface. Decrease all values of the thresholds (down - shift) will lead to more optimistic MCS assignments and hence, higher bitrates and possibly more HARQ retransmissions and higher BLERs. Keep the lower values unchanged and gradually increase/decrease the higher values. Such a modification will only force less/more robust MCSs (i.e. higher/lower data rates) for good propagation conditions. Keep the higher values unchanged and gradually increase/decrease the lower values. Such modifications will only force less/more robust MCSs (i.e. higher/lower data rates) for bad propagation conditions. Increase the lower values and decrease the higher values while keeping the middle values unchanged. Such a modification will force more robust MCSs (i.e. lower data rates) for bad propagation conditions and will force less robust MCSs (i.e. higher data rates) for good propagation conditions. Decrease the lower values and increase the higher values while keeping the middle values unchanged. Such modifications will force less robust modulations (i.e. higher data rates but possible higher BLER) for bad propagation conditions and more robust MCSs (i.e. lower data rates) for good propagation conditions. All types of thresholds tuning specified above can indeed be performed by changing the thresholds by various amounts. Finding the best type of modification and the amount by which the changes are made is part of the optimization process.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 75/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Throughput - proper tuning increases the throughput Capacity - proper tuning can increase the capacity (capacity reached over a slightly wider range of propagation conditions). For finding the optimum set of SINR thresholds among the sets proposed in the table above, a drive test is needed in the cell to be optimized. The flowing steps need to be performed and the optimum set of thresholds shall be chosen based on the observed performance. Step 1: In the eNodeB database, choose the default set of values for dlMCSTransitionTable, the values in the “Default” column in the table above. Step 2: While performing DL UDP transfer, perform a drive tests through the cell for covering various morphologies and positions relative to the transmitting antennas (near-cell, mid-cell, cell edge) and log the instantaneous throughput. Step 3: Chose another set of values from the table above and repeat Step 2.
10.1.2 DLSINRTHRESHOLDBETWEENCLMIMOONELAYERANDTXDIV dlSinrThresholdBetweenCLMimoOneLayerAndTxDiv defines the switching threshold between TxDiv and 1-layer CL-MIMO. Although the recommendation is to use the value -10 in order to keep as minimum the 1-layer CL-MIMO; disabling in this manner the Tx-Div.
Recommended & Default Value= "-10.0" Some considerations regarding this parameter: If parameter dlSinrThresholdBetweenCLMimoOneLayerAndTxDiv is set equal to dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer the downlink transmission scheme configured at cell level will be either TxDiv or 2-layer CLMIMO (i.e. CL-MIMO 1 layer is disabled), the switching threshold being dlSinrThresholdBetweenCLMimoOneLayerAndTxDiv = dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer in this case. If, besides, the reported (and filtered) rank is 1, the downlink transmission scheme will just be TxDiv. Also, if parameter macMIMOModeDl is set to “MimoTwoLayersNotAllowed”, the downlink transmission scheme configured at bearer-type level will be TxDiv for all the cells hosted by the eNB. Note that the configuration dlSinrThresholdBetweenCLMimoOneLayerAndTxDiv > dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer must be avoided. Increasing the value of this parameter would: Force TxDiv instead of 1-Layer Closed Loop which will be reflected in lower throughputs and performance for the same radio condition.
10.1.3 DLSINRTHRESHOLDBETWEENCLMIMOTWOLAYERSANDONELAYER
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 76/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer sets the SINR threshold for switching between the two transmission modes in transmission mode TM4 this is, sets the switching between the Closed Loop Mimo One Layer and Closed Loop Mimo Two Layers. High values will reduce Downlink data rate too soon.
Figure 10.1-3: Dl Sinr Threshold Example Recommended & Default Value= "12.0" Increasing the value of this parameter would: Force 1-Layer CL-MIMO in good transmission conditions which will be reflected in lower throughputs for good radio condition. Decreasing the value of this parameter would: Force 2-Layer CL-MIMO for bad propagation condition. This will be reflected in lower throughput at least for SINRs values situated between the actual value of the threshold and the optimized value of the threshold. Higher BLER and increased HARQ retransmissions might as well be observed. KPI Impact: Throughput - values both higher and lower than the optimal value decrease the throughput. For finding an optimized value of this parameter (i.e. the one that maximizes the throughput) for a given environment, a procedure containing the steps below can be used: Step 1: In the database, set transmission mode = tm4 and check that default value of dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer (i.e. 15) is also correctly set. Step 2: While performing DL FTP/UDP transfer, perform a drive tests through the cell and log the instantaneous throughput. Step 3: Repeat Step 1 & 2 for the following set of values of dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer: {9, 11, 13, and 17} Step 4: Provide results in form of a graph representing average throughput versus dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer values Step 5: Provide recommendations for choosing dlSinrThresholdBetweenCLMimoTwoLayersAndOneLayer
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 77/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-4: CL 2Layer-1Layer SNR Switch Threshold: 10 dB (purple) vs. 12 dB (blue) AWGN (Lab results VzW)
Figure 10.1-5: CL 2Layer-1Layer SNR Switch Threshold: 10 dB (purple) vs. 12 dB (blue) EPA 5Hz, Medium Correlation (Lab Results VzW)
10.1.4 DLSINRTHRESHOLDBETWEENOLMIMOANDTXDIV dlSinrThresholdBetweenOLMimoAndTxDiv Determines the value of the SINR to which there is a switch between the two transmission modes available in tm3 i.e. OL MIMO and Tx Div. Higher values will reduce DL data rate otherwise achievable in the higher SINR regime. Lower values would allow OL MIMO too soon, resulting in HARQ retransmission rates and BLERs higher than achievable with Tx diversity and hence the use of an MCS with a lower DL data rate/throughput. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 78/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-6: Dl Sinr Threshold Example Parameters dlSinrThresholdBetweenOLMimoAndTxDiv and dlSpeedThresholdBetweenOLMimoAndTxDiv configure thresholds ThSinrMimo and ThSpeedMimo, respectively.
Recommended & Default Value= "0.0"
NEA Recommended Value= “13.0” for density urban case. Actually, this value depends on radio environment and should be tuned according to deployed scenarios. Note that, the NEA recommended value is based on density urban scenario test results. Actually, the SINR threshold for intra transmission mode adaption should be different for different scenarios. The SINR threshold here is the CQI converted SINR which comes from cQItoSINRLookUpTable, it is not the CRS SINR. Increasing the value of this parameter would: Force TxDiv in good transmission conditions which will be reflected in lower throughputs for good radio condition. Decreasing the value of this parameter would: Force OL MIMO for bad propagation condition for which the MIMO algorithms are not anymore performing well. This will be reflected in lower throughput at least for SINRs values situated between the actual value of the threshold and the optimized value of the threshold. Higher BLER and increased HARQ retransmissions might as well be observed. KPI Impact: Throughput - values both higher and lower than the optimal value decrease the throughput.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 79/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
For finding an optimized value of this parameter (i.e. the one that maximizes the throughput) for a given environment, a procedure containing the steps below can be used: Step 1: In the database, set the transmission mode to tm3 and check that the default value of dlSinrThresholdBetweenOLMimoAndTxDiv (i.e. 15) is also correctly set. Step 2: While performing DL FTP/UDP transfer, perform 5 rounds drive test in the route path according to Figure 10.1-7 and log the instantaneous throughput, SINR, 2 codewords rate and DL BLER. Step 3: Repeat Step 1 and step 2 the following set of values of dlSinrThresholdBetweenOLMimoAndTxDiv: {10, 12 and 13} Step 4: Provide results in form of a graph representing average throughput versus dlSinrThresholdBetweenOLMimoAndTxDiv values Step 5: Provide recommendations for choosing dlSinrThresholdBetweenOLMimoAndTxDiv.
Figure 10.1-7: dl SINR Threshold for TM switching driver test route (CMCC LST urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 80/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
DL Phy throuthput(Mbps) & SINR
35.00
100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
30.00 25.00 20.00 15.00 10.00 5.00 0.00 15
13
12
Two codewords rate(%)
dlSinrThresholdBetweenOLMimoAndTxDiv vs. DL Phy throughput
SINR(dB) Phy thupt(Mbps) Two CWs rate(%)
10
dlSinrThresholdBetweenOLMimoAndTxDiv Figure 10.1-8: OL 2Layer-TxDiv SINR Switch Threshold vs Phy average throuthput (CMCC LST Density urban) Based on the above results in Figure 10.1-8, we can conclude that the average DL Physical layer throughput is the best while dlSINRThresholdbetweenOLMimoAndTxDiv is set to 13. The lower the threshold the higher the two code-words rate can get, but the average phy throughput does not have the same trend because TM3 still need suitable radio condition (SINR, RI>1) to reach 2layer MIMO, force 2 layer transmission cannot get benefit when SINR is low.
10.1.5 DLSINRTHRESHOLDBETWEENRANK1BEAMFORMINGANDTM3 TLA6.0 supports inter transmission mode auto-switch feature (FTS T115718), eNB has the capability to change the TM mode according to the channel condition by RRC reconfiguration. In TLA6.0, TM3/7 switching and TM3/8 switching are supported for 8 antennas. This switching is enabled when isInterTransmissionModeSwitchingEnabled is equal to “TM378switchingenaled”, it’s a license flag. Inter-mode switching is judged before intra-mode switching in L2 when adaptive TM mode is supported. Parameter dlSINRThresholdbetweenRank1BeamformingAndTM3 indicates the dl SINR threshold, which is used by DL scheduler to decide the transmission scheme switching decision between Rank1 DL BeamForming(TM7 beamforming or rank1 beamforming of TM8) and transmission mode 3. If UE is working on TM3 mode, UESpeed <= SpeedThresholdBetweenOLAndCL and SinrInterMimoEff <= dlSINRThresholdbetweenRank1BeamformingAndTM3 are all satisfied, ENB will signal UE to switch from TM3 to TM7. If UE is working on TM7 mode, and SinrInterMimoEff > dlSINRThresholdbetweenRank1BeamformingAndTM3 + DeltaSINR is satisfied. ENB will signal UE to switch to TM3. Recommended & Default Value= “5.0” Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 81/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
NEA Recommended Value= “10.0” for density urban. Actually, this value depends on radio environment and should be tuned according to deployed scenarios. Note that, the NEA recommended value is based on density urban scenario test results. Actually, the SINR threshold for inter transmission mode adaption should be different for different scenarios. The SINR threshold here is the CQI converted SINR which comes from cQItoSINRLookUpTable, it is not the CRS SINR.
Increasing the value of this parameter would: Force TM7 in good transmission conditions which will be reflected in lower throughputs for good radio condition. Decreasing the value of this parameter would: Force TM3 for bad propagation condition for which the MIMO algorithms are not anymore performing well. This will be reflected in lower throughput at least for SINRs values situated between the actual value of the threshold and the optimized value of the threshold. Higher BLER and increased HARQ retransmissions might as well be observed. For finding an optimized value of this parameter (i.e. the one that maximizes the throughput) for a given environment, a procedure containing the steps below can be used: Step 1: In the database, set the transmission mode to tm7 and check that the default value of dlSINRThresholdbetweenRank1BeamformingAndTM3 (i.e. 15) is also correctly set. Step 2: While performing DL full buffer FTP transfer, perform 3rounds drive test in the route path according to Figure 10.1-7 and log the instantaneous throughput, SINR, 2 codewords rate and DL BLER. Step 3: Repeat Step 1 and step 2 the following set of values of dlSINRThresholdbetweenRank1BeamformingAndTM3: {2, 3, 5, 7, 9, 10 and 12} Step 4: Provide results in form of a graph representing average throughput versus dlSINRThresholdbetweenRank1BeamformingAndTM3 values Step 5: Provide recommendations for choosing dlSINRThresholdbetweenRank1BeamformingAndTM3.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 82/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
PDCP Avg. Throuthput(Mbps)
dlSINRThresholdbetweenRank1BeamformingAndTM3 Vs. PDCP Average Throuthput 30.00 25.00 20.00 15.00
round1
10.00
round2 round3
5.00 0.00 12
10
9
7
5
3
2
dlSINRThresholdbetweenRank1BeamformingAndTM3 Figure 10.1-9: dlSINRThreshold for TM3/7 Switch vs. PDCP average throughput (CMCC LST Density urban)
Two codewords rate(%)
dlSINRThresholdbetweenRank1BeamformingAndTM3 Vs. Two codewords rate 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
round1 round2 round3
12
10
9
7
5
3
2
dlSINRThresholdbetweenRank1BeamformingAndTM3 Figure 10.1-10: dlSINRThreshold for TM3/7 Switch vs. 2 codewords rate (CMCC LST Density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 83/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
dlSINRThresholdbetweenRank1BeamformingAndTM3 Vs. TM7 rate 30%
TM7 rae(%)
25% 20% 15%
round1
10%
round2 round3
5% 0% 12
10
9
7
5
3
2
dlSINRThresholdbetweenRank1BeamformingAndTM3 Figure 10.1-11: dlSINRThreshold for TM3/7 Switch vs. TM7 rate (CMCC LST Density urban) Based on the above results in Figure 10.1-9 we can conclude that the average DL PDCP throughput is close and the best while dlSINRThresholdbetweenRank1BeamformingAndTM3 is set to 10 and 2. The same conclusion for two codewords rate can be gotten based on results in Figure 10.1-10. Based on the results in Figure 10.1-11 we can conclude that the lower the threshold the lower the TM7 rate can be used, which means UE is forced to use TM3 even when actual SINR is low. Actually, Beamforming have some gain compare to TxDiv when SINR is bad (in bad radio condition, cell edge, high interference, etc..), so we need UE to use Beamforming instead of TxDiv in this scenario, then the threshold should not be set to low value (e.g. 3, 2, etc..). Based on the above transmission scheme switching theory and test results, we suggest using value 10 in Density Urban scenario. Note: This parameter should be set cross link with parameter deltaSINRforIntermodeSwitch and dlSinrThresholdBetweenOLMimoAndTxDiv to improve DL performance.
10.1.6 DLSINRTHRESHOLDBETWEENRANK2BEAMFORMINGANDTM3 TLA6.0 supports inter transmission mode auto-switch feature (FTS T115718), eNB has the capability to change the TM mode according to the channel condition by RRC reconfiguration. In TLA6.0, TM3/7 switching and TM3/8 switching are supported for 8 antennas. This switching is enabled when isInterTransmissionModeSwitchingEnabled is equal to “TM378switchingenaled”, it’s a license flag. Inter-mode switching is judged before intra-mode switching in L2 when adaptive TM mode is supported. Parameter dlSINRThresholdbetweenRank2BeamformingAndTM3 indicates the dl SINR threshold, which is used by DL scheduler to decide the transmission scheme switching decision between DL dual layer BeamForming(TM8) and transmission mode 3. If UE is working on TM3 mode, UESpeed <= SpeedThresholdBetweenOLAndCL and Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 84/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SinrInterMimoEff <= dlSINRThresholdbetweenRank2BeamformingAndTM3 and UE can support TM8 (UE capability “enhancedDualLayerTDD-Supported”) and SRS SINR is enough to support BF are all satisfied, ENB will signal UE to switch from TM3 to TM8. If UE is working on TM8 mode, and SinrInterMimoEff > dlSINRThresholdbetweenRank2BeamformingAndTM3 + DeltaSINR and RankInterMimoEff> RankThresholdBetweenRank1AndRank2 or SinrInterMimoEff > dlSINRThresholdBetweenOLMimoAndTxDiv+hysteresisSinrInTM3 and RankInterMimoEff> RankThresholdBetweenRank1AndRank2 and UESpeed > SpeedThresholdBetweenOLAndCL + DeltaSpeed are satisfied. ENB will signal UE to switch to TM3.
Recommended & Default Value= “8.0”
NEA Recommended Value= “14.0” for density urban. Actually, this value depends on radio environment and should be tuned according to deployed scenarios. Note that, the NEA recommended value is based on density urban scenario test results. Actually, the SINR threshold for inter transmission mode adaption should be different for different scenarios. The SINR threshold here is the CQI converted SINR which comes from cQItoSINRLookUpTable, it is not the CRS SINR. Increasing the value of this parameter would: Force TM8 in good transmission conditions which will be reflected in lower throughputs for good radio condition. Decreasing the value of this parameter would: Force TM3 for bad propagation condition for which the MIMO algorithms are not anymore performing well. This will be reflected in lower throughput at least for SINRs values situated between the actual value of the threshold and the optimized value of the threshold. Higher BLER and increased HARQ retransmissions might as well be observed. For finding an optimized value of this parameter (i.e. the one that maximizes the throughput) for a given environment, a procedure containing the steps below can be used: Step 1: In the database, set the transmission mode to tm3/8 switch enabled (isInterTransmissionModeSwitchingEnabled= TM378switchingenabled ) and check that the default value of dlSINRThresholdbetweenRank2BeamformingAndTM3 (i.e. 8) is also correctly set. Step 2: While performing DL full buffer FTP transfer, perform 3rounds drive test in the route path according to Figure 10.1-7 and log the instantaneous throughput, SINR, 2 codewords rate and DL BLER. Step 3: Repeat Step 1 and step 2 the following set of values of dlSINRThresholdbetweenRank2BeamformingAndTM3: {6, 8, 9, 10, 12 and 14} Step 4: Provide results in form of a graph representing average throughput versus dlSINRThresholdbetweenRank2BeamformingAndTM3 values Step 5: Provide recommendations for choosing dlSINRThresholdbetweenRank2BeamformingAndTM3. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 85/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-12: Test road path for TM3/8 switching (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban)
The cross point of TM8 vs TM3
Figure 10.1-13: PDCP average throughput TM3 vs. TM8 (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 86/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-14: PDCP average throughput TM3 vs. TM3/8 switching (CMCC LTE TDD precommercial deployment Qingdao -Density urban) Based on the test result above in Figure 10.1-13, we can conclude that TM8 has performance gain at middle cell and cell edge than TM3, and then we can get benefit of switching between TM3 and TM8. That’s why we use TM3/8 inter-transmission mode switching to help to improve cell total throughput performance, as shown in Figure 10.1-14.
Average of DlPDCPThrough put_TM378_12+ 5
28-29
26-27
24-25
22-23
20-21
18-19
16-17
14-15
12-13
8-9
10-11
6-7
4-5
2-3
0-1
-2--1
-4--3
-6--5
-8--7
-10--9
-12--11
Average of DlPDCPThrough put_TM378_14+ 3
(blank)
70000 65000 60000 55000 50000 45000 40000 35000 30000 25000 20000 15000 10000 5000 0
-14--13
PDCP avg. Throughput
PDCP avg throughput vs CRS SINR @TM378 dlthreshold
CRS SINR
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 87/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-15: dlSINRthreshold for TM3/8 switching vs. PDCP average throughput (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban)
PDCP Avg. Throughput (Kbps)
(dlSINRThresholdbetweenRank2BeamformingAndTM3, deltaSINRforIntermodeSwitch) Vs. PDCP Avg. Throuthput 29500 29000 28500 28000 27500 27000
TM3/8switch(12,5)
26500
TM3/8switch(14,3)
26000 25500 TM3/8switch(12,5)
TM3/8switch(14,3)
(dlSINRThresholdbetweenRank2BeamformingAndTM3, deltaSINRforIntermodeSwitch) Figure 10.1-16: dlSINRthreshold for TM3/8 switching vs. PDCP average throughput (CMCC LTE TDD pre-commercial deployment Qingdao -Density urban) According to the test results in Figure 10.1-15 and Figure 10.1-16, we found that the optimized values of (dlSINRThresholdbetweenRank2BeamformingAndTM3, deltaSINRforIntermodeSwitch) combination setting could be (14, 3) in the test scenario. Actually, the fine tuning of the combination setting of (dlSINRThresholdbetweenRank2BeamformingAndTM3, deltaSINRforIntermodeSwitch) depends on radio environment (results as CQI, RI) and should be tuned according to deployed scenarios. That is, different scenarios may have different radio environments and shall use different threshold for inter transmission mode switch accordingly to enlarge the benefit of inter-mode switch.
10.1.7 DELTASINRFORINTERMODESWITCH As described in section 10.1.5&10.1.6, deltaSINRforIntermodeSwitch is another key parameter for inter transmission mode switch. This parameter is the hysteresis of inter transmission mode switching, which is used to avoid the ping-pong switching of 2 transmission mode.
Recommended & Default Value= “3.0”
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 88/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Increasing the value of this parameter would: Make it harder to switch from TM8 to TM3 and less ping-pong switching occurs. High value may force TM8 even in good transmission conditions, which will be reflected in lower throughputs for good radio condition. Decreasing the value of this parameter would: Make it easier to switch from TM8 to TM3 and ping-pong switching occurs frequently. Low value may force TM3 for bad propagation condition for which the MIMO algorithms are not anymore performing well. This will be reflected in lower throughput at least for SINRs values situated between the actual value of the threshold and the optimized value of the threshold. Higher BLER and increased HARQ retransmissions might as well be observed. This parameter shall be optimized together with parameter dlSINRThresholdbetweenRank2BeamformingAndTM3 for TM3/8 switching and parameter dlSINRThresholdbetweenRank1BeamformingAndTM3 for TM3/7 switching, respectively. For TLA6.0 field test results of combination tuning of (dlSINRThresholdbetweenRank2BeamformingAndTM3, deltaSINRforIntermodeSwitch), please refer to section 10.1.6.
10.1.8 DOWNLINK PERFORMANCE COMPARISON OF DIFFERENT TRANSMISSION MODE Because TLA6.0 supports TM1/2/3/7/8 when using 8 antennas, we did many tests to compare the downlink performance between TM2, TM3, TM7 and TM8 in the same environment with same configuration and same UE. One test environment is CMCC LST density urban scenario with LOS, we selected cell1 of site Lvkai in CMCC LST south pudong road area, according to Figure 10.1-17; the other is Jinqiao suburban scenario with few LOS, we selected cell3 of site Jinqiao OTA in Jinqiao, according to Figure 10.1-18. For comparing the performance of different transmission for a given environment, a procedure containing the steps below can be used: Step 1: Set the value of isInterTransmissionModeSwitchingEnabled to “disabled” which disables inter transmission mode switch function, other parameters use the default value except the parameters in the following steps. Step 2: Test TM2 performance. In the database, set the transmission mode to tm2. Step 3: While performing DL full buffer FTP transfer, perform 5 rounds drive test with speed=<30Km/h in the route paths according to Figure 10.1-17 and Figure 10.1-18 by using Hisilicon R8 driver test UE and Innofidei R9 driver test UE, respectively, and log the instantaneous throughput, SINR, CQI, MCS, 2 codewords rate and DL BLER. Step 4: Test TM3 performance. In the database, set the transmission mode to tm3. Check that the value of dlSINRThresholdbetweenRank1BeamformingAndTM3 is set to -10 to enforce UE use TM3 with wireless environment ignored, then repeat step 3. Step 5: Test TM7. In the database, set the transmission mode to tm7. Check that the value of uLCESINRThresholdBetweenTxDivAndBeamFormingIntraTm7 is set to -7 to enforce UE use TM7 with wireless environment ignored, then repeat step 3. Step 6: Test TM8. In the database, set the transmission mode to tm8. Check that the value of dlSinrThresholdBetweenRank1BeamFormingAndRank2BeamForming is set to 5, dlSinrThresholdBetweenTxDivAndRank1BeamForming is set to 0 and RIThresholdBetweenRank1AndRank2 is set to 0.6, then repeat step 3. Step 7: Provide results in form of a graph representing average throughput versus CRS SINR. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 89/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-17: Test route path for TM performance comparison (site Lvkai in CMCC LST density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 90/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 10.1-18: Test route path for TM performance comparison (OTA in Jinqiao_Suburban)
10.1.8.1 PERFORMANCE COMPARISON IN CMCC LST DENSITYURBAN SCENARIO
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 91/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR-Bin vs. Tput -average (30km/h) 60 Tput
SPuDong-Inno-TM2 SPuDong-Hisi-TM2
50
SPuDong-Inno-TM3 SPuDong-Hisi-TM3
40
SPuDong-Inno-TM7
30
SPuDong-Hisi-TM7 SPuDong-Inno-TM8
20 10 0 -10
-5
0
5
10
15
20
25
30
35
SINR
-10
Figure 10.1-19: SINR vs. Avg. Throughput of TM2/3/7/8 (site Lvkai in CMCC LST density urban)
SINR vs. MCS (30km/h) MCS 30
SPuDong-Inno-TM2 SPuDong-Hisi-TM2 SPuDong-Inno-TM3
25
SPuDong-Hisi-TM3 SPuDong-Inno-TM7
20
SPuDong-Hisi-TM7
MCS turn point
15
sPuDong-Inno-TM8
10
5 SINR
0
-10 -5 0 5 10 15 20 25 30 Figure 10.1-20: SINR vs. MCS of TM2/3/7/8 (site Lvkai in CMCC LST density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 92/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR vs. BLER (30km/h) SPuDong-Inno-TM2
BLER 50
SPuDong-Hisi-TM2
45
SPuDong-Inno-TM3
40
SPuDong-Hisi-TM3
35
Bler increase while MCS increase
30 25
SPuDong-Inno-TM7 SPuDong-Hisi-TM7 sPuDong-Inno-TM8
20
15 10 5 SINR
0
-10 -5 0 5 10 15 20 25 30 Figure 10.1-21: SINR vs. BLER of TM2/3/7/8 (site Lvkai in CMCC LST density urban)
SINR vs. CQI0 (30km/h) SPuDong-Inno-TM2
CQI0 16
SPuDong-Hisi-TM2
14
SPuDong-Inno-TM3 SPuDong-Hisi-TM3
12
SPuDong-Inno-TM7
10
SPuDong-Hisi-TM7
CQI cross points
8
sPuDong-Inno-TM8
6 4 2 SINR
0
-10 -5 0 5 10 15 20 25 30 Figure 10.1-22: SINR vs. CQI_0 of TM2/3/7/8 (site Lvkai in CMCC LST density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 93/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Based on the above results in Figure 10.1-19 we can conclude that the average DL throughput has the trend as below: For Hisilicon R8 UE, TM3>TM2>TM7@SINR>2dB in the mass, while TM3 and TM2 had 2 cross points: one is at SINR=11dB, the other is at SINR=9dB, but TM3 and TM2 are close when SINR=<11dB. We can conclude that TM3/Txdiv threshold can be around SINR=11dB at this scenario. TM7 only shows tiny gain than TM3/TM2 when SINR <2dB. For Innofidei R9 UE, TM3>TM8 at good coverage (SINR>17dB) and TM3>TM2>TM7@SINR>14dB, the first cross point of TM3 and TM8 is at SINR=17dB, and TM3 throughput is close to TM8 when SINR belong to [3,17]dB; TM8 shows small gain than TM3 when SINR<3dB, it also shows small gain than TM2 when SINR<8dB;TM7 shows some gain than TM3, TM2 and even TM8 when SINR<13dB, but close to TM2 and TM8 when SINR=<0dB; TM2 shows small gain than TM3 and close to TM8 when SINR <3dB.
Based on the above results in Figure 10.1-20 and Figure 10.1-21 we can conclude that the MCS turn point indicator the 2 layer to 1 layer changes for TM3 and TM8, the MCS increases lead to Bler increasing and bring negative impacts on DL throughput. Based on the above results in Figure 10.1-22 we can conclude that the TM3 codeword0 CQI of Hisilicon UE and Innofidei UE is quite close when at middle cell (e.g. SINR >=8dB), and the TM2/TM7 CQI values of the 2 type UEs are quite same with same SINR value, which means the radio environment for test is stable and the same, and 1 layer CQI measurement results of these 2 type of UEs are close. Based on Figure 10.1-22, we also found that when Hisilicon R8 UE use TM3 in this scenario, the codeword0 CQI is almost the same when SINR is bad (=<6dB), which is unreasonable, meanwhile, the MCS and throughput of TM3 are close to TM2 according to Figure 10.1-19 and Figure 10.1-20, which is as what we expected, so, maybe the CQI values we got from UE when SINR=<6dB was error. On the other hand, the behaviours of the 2 type of UEs are different in this scenario: For Hisilicon R8 UE, its’ TM3 has the best performance, but its’ TM7 performance is the worst; For Innofidei R9 UE, its’TM3 and TM2 performance are not as good as Hisilicon, but its’ TM7 performance is better than Hisilicon, even than its’ TM8, and its’ TM8 performance is worse than Hisilicon’s TM3 when SINR>3dB, but close to its’ TM3 and Hisilicon’s TM2 when SINR<17dB.
To conclude, in this density urban with LOS test scenario, the DL performance trend can be as TM3>TM8>TM2>TM7 when SINR is not too bad (e.g. SINR>=3dB), TM8=TM7 and shows small gain than TM3 & TM2 when at cell edge (e.g. SINR<3dB).
10.1.8.2 PERFORMANCE COMPARISON IN JINQIAO OTA SUBURBAN SCENARIO
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 94/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR-Bin vs. Tput -average (30km/h) 60 Tput
TM2-Inno
50
TM2-Hisi
40
TM3-Inno-Threshold10dB TM3-Hisi-Threshold10dB TM7-Inno-30km/h
30
TM7-Hisi-30km/h
20
TM8-Inno-30km/h
10 0 -20
-10
0
10
20
30
-10
40 SINR
Figure 10.1-23: SINR vs. Avg. Throughput of TM2/3/7/8 (OTA in Jinqiao suburban)
SINR vs. MCS (30km/h)
TM2-Inno
MCS 30
TM2-Hisi TM3-Inno-Threshold-10
25
TM3-Hisi-Threshold10dB TM7-Inno-30km/h
20
TM7-Hisi-30km/h
15
TM8-Inno-30km/h
MCS turn point
10
5 SINR 0 -10 -5 0 5 10 15 20 25 30 Figure 10.1-24: SINR vs. MCS of TM2/3/7/8 (OTA in Jinqiao suburban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 95/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR vs. BLER (30km/h)
TM2-Inno
BLER 45
TM2-Hisi
40
TM3-InnoThreshold-10dB TM3-Hisi-Threshold10dB TM7-Inno-30km/h
Bler increase while MCS increase
35 30 25
TM7-Hisi-30km/h
20
TM8-Inno-30km/h
15 10 5 0 -10
-5
0
5
10
15
20
25
30
SINR
Figure 10.1-25: SINR vs. BLER of TM2/3/7/8 (OTA in Jinqiao suburban)
SINR vs. CQI0 (30km/h)
TM2-Inno
CQI0 16
TM2-Hisi
14
Tm3-InnoThreshold-10dB TM3-HisiThreshold-10dB TM7-Inno-30km/h
12 10 8
CQI cross point
6
TM7-Hisi-30km/h TM8-Inno-30km/h
4 2 0 -10
-5
0
5
10
15
20
25
30
SINR
Figure 10.1-26: SINR vs. CQI_0 of TM2/3/7/8 (OTA in Jinqiao suburban) Based on the above results in Figure 10.1-23 we can conclude that the average DL throughput has the trend as below: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 96/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
For Hisilicon R8 UE, TM3>TM2 >TM7@SINR>7dB, actually, TM3 and TM2 had one cross point at SINR=3dB, i.e. TM3>TM2 until SINR=3dB, after that, TM3 and TM2 are almost the same; TM7 shows small gain than TM3/TM2 when SINR <7dB, we can conclude that the threshold between TM7 and TM3 can be 7dB at this scenario. For Innofidei R9 UE, TM3>TM2>TM7@SINR>7dB, actually, the first cross point of TM3 and TM2 is at SINR=13dB, i.e. TM3>TM2 when SINR>13dB, the 2nd cross point of TM3 and TM2 is at SINR 9dB, i.e. TM2>TM3 when SINR belong to [9,13] dB, after that TM3 and TM2 are almost the same; TM7 shows small gain than TM3/TM2 when SINR<7dB; TM8 only got a bit higher throughput than TM7 when SINR>19dB, after SINR<12dB, TM8 performance is lower than TM7.
Based on the above results in Figure 10.1-24 and Figure 10.1-25 we can also conclude that the MCS increase points indicator the 2 layer to 1 layer changes for TM3, the MCS increases lead to Bler increasing and bring negative impacts on DL throughput. Based on the above results in Figure 10.1-26 we can conclude that the TM3 codeword0 CQI of Hisilicon UE and Innofidei UE is quite close when SINR>=13dB; the first cross point of TM3 and TM2/TM7 is SINR=5dB for Hisilicon R8 UE, and SINR=3dB for Innofidei R9 UE; TM2/TM7 CQI values of the 2 type UEs are also quite same with same SINR value, which means the radio environment for test is stable and the same, and 1 layer CQI measurement results of these 2 type of UEs are close. Based on Figure 10.1-26, we found that when Hisilicon R8 UE use TM3 in this scenario, the lower the SINR the lower the codeword0 CQI, which is as what we expected, the CQI values error issue occurred in site Lvkai did not occur anymore. Innofidei R9 UE TM8 only shows higher performance than its’ TM7 when SINR>19dB and Hisilicon R8 UE TM7 when SINR>15dB, TM8 also shows higher performance than TM2 when SINR>25dB, besides that TM8 shows no benefit than any other transmission mode, which is not as what we expected. Based on Figure 10.1-24 and Figure 10.1-26, we found that this is because Innofidei R9 UE TM8 got lower CQI and MCS than any other transmission mode in Jinqiao OTA scenario. On the other hand, the behaviours of the 2 type of UEs are different in this scenario: For Hisilicon R8 UE, its’ TM3 still has the best performance, and its’ TM7 performance is better than Innofidei when SINR <8dB; For Innofidei R9 UE, its’TM3 performance is not as good as Hisilicon, but its’ TM7 performance is better than Hisilicon when SINR>=16dB, and its’ TM8 did not get the position as we expected. Both 2 type of UEs have almost the same TM2 performance in this scenario, and TM7 has small gain than TM3/TM2 when SINR<7dB, TM8 shows no benefit when SINR<15dB.
To conclude, in this suburban with few LOS test scenario, the DL performance trend can be as TM3>TM8> TM2>TM7 when in near cell (e.g. SINR>25dB), TM3>TM2>TM8>TM7 when in good radio condition (e.g. SINR>19dB) and TM7>TM3/TM2>TM8 when SINR is not good (e.g. <7dB) (actually, TM8 performance is lower than others when SINR<15dB). TM7 shows small gain than TM3 & TM2 when SINR<7dB but TM8 shows no gain than others when SINR<15dB, the TM8 performance is not as good as what we expected, which may due to the TM8 of Innofidei R9 UE is not well optimized, another impact factor is the less LOS paths in Jinqiao OTA scenario.
10.1.8.3
PERFORMANCE COMPARISON OF BOTH SCENARIOS
TM3 performance comparison in both scenarios
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 97/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR-Bin vs. Tput -average (30km/h)
60 Tput
OTA-Inno-TM3
50
OTA-Hisi-TM3
40
SPuDong-Inno-TM3
30
SPuDong-Hisi-TM3
20 10 0 SINR -10 -5 0 5 10 15 20 25 30 35 Figure 10.1-27: SINR vs. Avg. Throughput of TM3 in both scenarios
MCS
30
SINR vs. MCS (30km/h)
OTA-Inno-TM3 OTA-Hisi-TM3
25
SPuDong-Inno-TM3
20
SPuDong-Hisi-TM3
15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-28: SINR vs. MCS of TM3 in both scenarios
25
30
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 98/290
LTE Optimization Handbook TLA6.0
BLER
45
MGR/TIPS/NEA
SINR vs. BLER (30km/h)
OTA-Inno-TM3
40
OTA-Hisi-TM3
35 SPuDong-Inno-TM3
30
SPuDong-Hisi-TM3
25 20 15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-29: SINR vs. BLER of TM3 in both scenarios
CQI0
16
25
30
SINR vs. CQI0 (30km/h)
OTA-Inno-TM3
14
OTA-Hisi-TM3
12
SPuDong-Inno-TM3
10
SPuDong-Hisi-TM3
8 6 4 2 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-30: SINR vs. CQI_0 of TM3 in both scenarios
25
30
Based on the above results we can conclude that both 2 type of UEs have better DL throughput performance in site Lvkai scenario than in site Jinqiao OTA scenario. Although UE can early report higher codeword0 CQI in OTA scenario, but the MCS UE can get is lower than in site Lvkai scenario, because the MCS changing occurred early in site Lvkai scenario, thus eNB indicator UE use 1 layer instead of 2 layer in site Lvkai scenario although UE still report RI>1, which means here exists mismatch in TM3 transmission. And the higher Bler in OTA scenario did negative impact on throughput while UE still using 2 layer transmission. So, we can find that the transmission mode mismatch scheme helped to improve throughput in this scenario with LOS, because MIMO is more suitable for more multi-paths radio scenario. But the different scenarios may get different results. TM8 performance comparison in both scenarios Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 99/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR-Bin vs. Tput -average (30km/h) Tput
45 40
SPuDong-InnoTM8
35
OTA-Inno-TM8
30 25 20 15 10 5 0 -20
-10
-5
0
10
20
40 SINR
30
Figure 10.1-31: SINR vs. Avg. Throughput of TM8 in both scenarios
SINR vs. MCS (30km/h) MCS 25
SPuDong-InnoTM8 OTA-Inno-TM8
20
15
MCS turn point
10
5 SINR 0 -10 -5 0 5 10 15 20 Figure 10.1-32: SINR vs. MCS of TM8 in both scenarios
25
30
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 100/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR vs. BLER (30km/h) BLER 45
SPuDong-InnoTM8
40
OTA-Inno-TM8
35 30 Higher Bler in OTA scenario
25 20 15 10 5
SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-33: SINR vs. BLER of TM8 in both scenarios
25
30
SINR vs. CQI0 (30km/h) CQI0 16
SPuDong-InnoTM8
14
OTA-Inno-TM8
12 10 8
6
CQI turn point
4
2 SINR 0 -10 -5 0 5 10 15 20 25 30 Figure 10.1-34: SINR vs. CQI_0 of TM8 in both scenarios Based on the above results we can conclude that Innofidei R9 UE has better DL throughput performance in site Lvkai scenario than in site Jinqiao OTA scenario.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 101/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Innofidei R9 UE reported lower codeword0 CQI in Jinqiao OTA scenario than in site Lvkai scenario during the test, especially when SINR<15dB, and the MCS UE can get in Jinqiao OTA scenario had the same trend. And based on results in Figure 10.1-32, we found that there was no MCS increasing occurred although there also had CQI turn point when SINR<11dB in Jinqiao OTA, which means the UE reported CQI may have changed to 1layer CQI, but there was always 2 codewords HARQ Bler, which means eNB still scheduled UE as 2layer and no mismatch for TM8 occurred. But UE behaviour in site Lvkai is different, the MCS changing occurred and eNB indicator UE use 1 layer instead of 2 layer in site Lvkai scenario although UE may still report RI>1 when SINR<15dB, which means here exists mismatch in TM8 transmission. The different behaviour of UE in the two different scenario may caused by UE TM8 capability is not well optimized and the LOS paths condition. And the higher Bler in Jinqiao OTA scenario did negative impact on throughput while UE still using 2 layer transmission. Based on the different behaviour in these two scenarios, we can conclude that TM8 should also get better performance when there are lots of LOS paths in scenario with LOS. TM2 performance comparison in both scenarios
Tput
35
SINR-Bin vs. Tput -average (30km/h) OTA-Inno-TM2
30
OTA-Hisi-TM2
25
SPuDong-Inno-TM2
20
SPuDong-Hisi-TM2
15 10 5 0
-10
-5
-5
0
5
10
15
20
25
30
35
SINR
Figure 10.1-35: SINR vs. Avg. Throughput of TM2 in both scenarios
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 102/290
LTE Optimization Handbook TLA6.0
MCS
30
MGR/TIPS/NEA
SINR vs. MCS (30km/h)
OTA-Inno-TM2 OTA-Hisi-TM2
25
SPuDong-Inno-TM2
20
SPuDong-Hisi-TM2
15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-36: SINR vs. MCS of TM2 in both scenarios
BLER
45
25
30
SINR vs. BLER (30km/h)
OTA-Inno-TM2
40
OTA-Hisi-TM2
35 SPuDong-Inno-TM2
30
SPuDong-Hisi-TM2
25 20 15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-37: SINR vs. BLER of TM2 in both scenarios
25
30
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 103/290
LTE Optimization Handbook TLA6.0
CQI0
MGR/TIPS/NEA
SINR vs. CQI0 (30km/h)
16
OTA-Inno-TM2
14
OTA-Hisi-TM2
12
SPuDong-Inno-TM2
10
SPuDong-Hisi-TM2
8 6 4 2 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-38: SINR vs. CQI_0 of TM2 in both scenarios
25
30
Based on the above results we can conclude that both 2 type of UEs have almost the same CQI report, the very small performance difference depends on MCS and Bler difference between the two test scenarios. Actually, we can find that TM2 does not very sensitive to the radio environment. TM7 performance comparison in both scenarios
SINR-Bin vs. Tput -average (30km/h)
30 Tput
OTA-Inno-TM7
25
OTA-Hisi-TM7
20
SPuDong-Inno-TM7 SPuDong-Hisi-TM7
15 10 5 0 -10
-5
-5
0
5
10
15
20
25
30 SINR
Figure 10.1-39: SINR vs. Avg. Throughput of TM7 in both scenarios
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 104/290
LTE Optimization Handbook TLA6.0
MCS
30
MGR/TIPS/NEA
SINR vs. MCS (30km/h)
OTA-Inno-TM7 OTA-Hisi-TM7
25
SPuDong-Inno-TM7
20 SPuDong-Hisi-TM7
15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-40: SINR vs. MCS of TM7 in both scenarios
BLER
45
25
30
SINR vs. BLER (30km/h)
OTA-Inno-TM7
40
OTA-Hisi-TM7
35
SPuDong-Inno-TM7
30
SPuDong-Hisi-TM7
25 20 15 10 5 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-41: SINR vs. BLER of TM7 in both scenarios
25
30
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 105/290
LTE Optimization Handbook TLA6.0
CQI0
16
MGR/TIPS/NEA
SINR vs. CQI0 (30km/h)
OTA-Inno-TM7
14
OTA-Hisi-TM7
12
SPuDong-Inno-TM7
10
SPuDong-Hisi-TM7
8 6 4 2 SINR
0 -10 -5 0 5 10 15 20 Figure 10.1-42: SINR vs. CQI_0 of TM7 in both scenarios
25
30
Based on the above results we can conclude that both 2 type of UEs have almost the same CQI report, but they did not have the same behaviour: Innofidei R9 UE got higher MCS when SINR>3dB, and lower Bler when SINR> 19dB in site Lvkai scenario than in OTA scenario, and got higher throughput performance in site Lvkai scenario when SINR>-1dB. Hisilicon R8 UE got higher MCS when SINR>3dB, and lower Bler when SINR> 11dB in site Lvkai scenario than in OTA scenario, but its’ throughput performance is higher in site Lvkai scenario only when SINR>9dB, after that, its’ throughput is worse in site Lvkai scenario.
Actually, beamforming needs to get DOA of UE to calculate the weight for visual beam, it should get better performance when there are lots of LOS paths, and this is verified by Innofidei R9 UE TM7 performance test comparison between the two test scenarios. Although the TM7 performance comparison of Hisilicon R8 UE did not completely show the same conclusion, but we think this is because the TM7 performance of Hisiliocn R8 UE does not be well optimized. In the summary, from the comparison we can conclude that in both two scenarios, TM3 has the best performance and has the trend that TM3>TM2>TM7 when SINR is higher than a threshold, but this threshold is not the same for different scenarios (e.g. threshold is 3dB for site Lvkai, and 7dB for Jinqiao OTA). As 3GPP defined, Txdiv is included in TM3, TM7 and TM8, so, for inter transmission mode switch, TxDiv will be used to transmit data while in HARQ hold state during the inter-mode switching, however, Txdiv in TM7 or TM8 uses resource allocation type2 which will limit the available RB number and lead to low throughput.. Furthermore, inter-mode switch is sensitive to channel environment, we should also consider the additional overhead and latency of signalling for inter transmission mode switching, the actual performance need to be estimated.
10.1.9 ALPHAFAIRNESSFACTOR AlphaFairnessFactor: tunes the alpha fairness factor of the DL scheduler alphaFairnessFactor = 0 yields a maximum C/I scheduler. The scheduler provides more resources to UEs in better conditions Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 106/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
alphaFairnessFactor = 1 yields a fair scheduler. The scheduler attempts to provide the same number of RBs to all the UEs alphaFairnessFactor = 2 yields an increased fairness scheduler. The scheduler attempts to allocate the resources in such a way that all the UEs eventually get the same data rate.
Deafult Value= "1" AlphaFairnessFactor is tested with the following procedure: Step Step Step Step Step Step
1: ENB configuration:20M, sa1+ssp7, TM2, and uplinkControlChannelLUTindex=1 2: the UEs’ distribution is depicted in Figure 10.1-43 below 3: set alphaFairnessFactor = 0 4: All UEs are attached and perform fullbuffer downlink FTP transmission, the related data are recorded. 5: All UEs are attached and perform 3M (or 6M) downlink UDP transmission, the related data are recorded. 6: reconfigure alphaFairnessFactor with the value 1 and 2, repeat step 4&5 respectively.
Figure 10.1-43: UEs’ distribution illustration In the Figure 10.1-44 it is presented the impact of using the different values for alphaFairnessFactor; such as 0 or 2.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 107/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Throughput (Kbps)
Throughput Vs. alpha fairness factor 4,700 4,200 3,700 3,200 2,700 2,200 1,700 1,200 700 200 -300
32,000 30,000 28,000 26,000 24,000 22,000 20,000 Max C/I UE #01 UE #07 UE #13
UE #02 UE #08 UE #14
PF UE #03 UE #09 UE #15
18,000
Enhanced PF UE #04 UE #10 Total
UE #05 UE #11
UE #06 UE #12
Figure 10.1-44: alphaFairnessFactor Change Impact with UE in different condition The test in Figure 10.1-44 was performed with the usual configuration; 5NC; 5MC; 5CE, all UEs performed full buffer DL FTP traffic and having the different alphafairnessFactor will make the scheduler to adjust more or less fair the resources to all the ue's; resulting in a lower sector t-put for the more fair distribution of resources. In the Figure 10.1-45 Impact of using different values of alpahFairnessFactor is depicted with different traffic mode (1, all UEs used 3Mbps DL FTP traffic; 2, all UEs used 6Mbps DL FTP traffic).
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 108/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Throughput (Kbps)
Throughput Vs. AlphaFairnessFactor 4,200 3,700 3,200
32,000
2,700
30,000
2,200 1,700
29,000
31,000
1,200 700 200 -300
28,000 27,000 3M
6M
3M
PF UE #01 UE #07 UE #13
6M Max C/I
UE #02 UE #08 UE #14
UE #03 UE #09 UE #15
3M
26,000
6M
enhanced PF UE #04 UE #10 Total
UE #05 UE #11
UE #06 UE #12
Figure 10.1-45: alphaFairnessFactor Change Impact – further details The test in Figure 10.1-45 was performed with different DL traffic mode; still 5NC; 5MC; 5CE, 1st round all UEs performed 3Mbps DL UDP traffic, 2nd round all UEs performed 6Mbps DL UDP traffic; two rounds have the different alphafairnessFactor will make the scheduler to adjust more or less fair the resources to all the ue's; resulting in a higher sector t-put for the more DL t-put requirement in the same fairness factor and a lower sector t-put for the more fair distribution of resources in the same DL traffic mode. KPI Impact: Throughput – Low values increase the throughput in the near&mid-cell condition at expense of cell edge users. Capacity - Low values increase the cell overall throughput at expense of cell edge users.
10.1.10 DYNAMICCFIENABLED This parameter when set to “True” allows the CFI to be dynamically adjusted to use the lowest value needed for PDCCH usage. This makes more OFDM symbols available to PDSCH when PDCCH usage is low (fewer users), resulting in higher throughputs. In this case (dynamicCFIEnabled set to “True), parameter cFI is ignored. When set to “False”, the CFI is static and derived from parameter cFI. The latter should be set keeping in mind that value “1” is only supported in 20 MHz and knowing that higher values of CFI allow for more PDCCH robustness and/or more users served per TTI, but at the expense of throughput (fewer resources for PDSCH). Recommended & Default Value= “false”
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 109/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
In the Table 10-2 it is illustrated the theory impact on the PDSCH channel when disabling the DynamicCFI and “forcing” the usage of different values for CFI (1, 2 or 3).
OFDM symbol per 1ms
CFI
PDSCH [symbol]
PDSCH usage[%]
PDSCH decrease [%]
14
1
13
92,86%
n/a
14
2
12
85,71%
7,14%
14
3
11
78,57%
7,14%
Table 10-2: Theory Assumption on CFI Tuning
10.1.11 CFI This parameter “Control Format Indicator” is limited to the value 1, 2 or 3. For bandwidths greater than ten resource blocks, the number of OFDM symbols used to contain the downlink control information is the same as the actual CFI value. Otherwise span of the downlink control information is CFI+1 symbol. For TDD Subframe 1 and 6, CFI is limited to the value1 or 2. It will be only taken into consideration if the dynamicCFIEnabled is set to “False”. Higher values of CFI allow for more PDCCH robustness and/or more users served per TTI, but at the expense of throughput (fewer resources for PDSCH). In the Table 10-2 we have the expected value when “playing” with the value of CFI. Recommended & Default Value= “3” for both 10MHz and 20MHz in trial mode.
KPI Impact: Throughput – Low values increase the cell overall throughput.
Note: For a demo case & if you are in 20MHz and the goal is to show max troughput you can use the CFI set to “1”; but to be sure that also dynamicCFIEnabled is set to “false”; plus CFI1Llowed should set to "True". Using this set of parameters configuration and in near cell radio condition (optimum conditions); you can boost the Downlink Throughput, since this makes more OFDM symbols available to PDSCH when PDCCH usage is low (fewer users), resulting in higher throughputs.
10.1.12 CFI1ALLOWED Like the name indicates this parameter allows the usage or not of the value “1” for CFI. Only possible to use CFI =1 in 20MHz. Recommended & Default Value= “10MHz = false for 20MHz = true”
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 110/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
10.1.13 CFI2ALLOWED Like the name indicates this parameter allows the usage or not of the value “2” for CFI. Recommended Value & Default= "True"
10.1.14 CFI3ALLOWED Like the name indicates this parameter allows the usage or not of the value “3” for CFI. Recommended Value & Default= "True"
10.1.15 MEASUREMENT GAP Measurement Gap is used to allow UE implement measurement for inter-frequency or inter-RAT measurement Mobility. The Measurements Gaps are small and periodic "gaps" in transmission, during which there is no DL transmission between the eNB and UE, and no UL transmission between this UE and the eNB. The length and the period of the gaps are determined by the eNB. There are 2 Gap Pattern Configurations:
Gap Pattern Id
MeasurementGap Length (MGL, ms)
Measurement Gap Repetition Period (MGRP, ms)
0 6 40 6 80 1 Table 10-3: Measurement Gap Pattern configurations
Measurement Gap Offset (MGO, ms)
0..39 0..79
According to 3GPP 36.133, it is defined that: During the measurement gaps the UE: shall not transmit any data is not expected to tune its receiver on the E-UTRAN serving carrier frequency. In the uplink subframe occurring immediately after the measurement gap, the E-UTRAN FDD UE shall not transmit any data the E-UTRAN TDD UE shall not transmit any data if the subframe occurring immediately before the measurement gap is a downlink subframe.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 111/290
LTE Optimization Handbook TLA6.0
0 D
1 S
2 U
SFN mod 2 = 0 4 5 6 D D S
3 U
7 U
8 U
9 D
MGR/TIPS/NEA
1 S
0 D
2 U
3 U
SFN mod 2 = 1 4 5 6 D D S
7 U
8 U
9 D
0 D
1 S
2 U
3 U
SFN mod 2 = 0 4 5 6 D D S
7 U
8 U
For E-UTRAN TDD, the UE shall not transmit in the uplink subframe occurring immediately after the measurement gap if the subframe occurring immediately before the measurement gap is a downlink subframe.
UL Dynamic
0
1
2
3
4
5
6
Offset between UL and DL subframes DL Dynaic 0
1
2
3
4
5
The unavailable subframe for UL or DL respectively
Measurement Gap
The unavailable dynamic scheduling subframe before MG due to without ACK/NACK feedback
The available subframe for UL or DL VoIP
The unavailable dynamic schedulingsubframe after MG for UL due to without UL grant only when nonSPS is configurated
Figure 10.1-46: Measurement Gap on DL and UL transmissions for TDD So, with Measurement Gap enabled, all DL/UL transmission is disabled: CQI reporting on PUCCH if active (UL)
CQI reporting on PUSCH if active (UL)
RI reports (UL)
SRS (UL)
Scheduling grants on PDCCH, except for RACH and Paging (DL)
Ack/Nack events on PHICH (DL) and PUCCH (UL) (except the last Ack/Nack when the max number of retransmissions for the process has been reached)
VoIP Semi-persistent scheduled events if active (UL & DL)
D-BCH pre-booked events on PDSCH (SIB1, SIB2 to SIB8), except 1.4MHz (due to the small PRB resource in 1.4MHz) (DL)
SR events (UL)
HARQ retransmission on DL/UL
The main points that drive the Measurement Gaps support are:
MG shall degrade UE performance as little as possible
Total cell throughput shall not been degraded at full cell capacity when MG are active for some or all the UEs.
MG is supported for all UEs, in all configurations.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 112/290
9 D
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Default Value= “length6ms_period40ms”.
NEA Recommended Value= “length6ms_period80ms”.
KPI Impact: Throughput – MG will degrade UE throughput. The shorter the Gap period the more the UE throughput degrade. Latency - long Gap length increase the latency. We collected field test results in CMCC Qingdao LTE TDD pre-commercial deployment, where we implemented driver test at NC of site SFChangshaRoad in cluster #16, and compared DL performance with/without MG enabled and with 40ms/80ms MG pattern. The TDD UL/DL configuration during the test is cfg2/7 with 20MHz BW.
Figure 10.1-47: Test road path for MG comparison The impact of MG on DL&UL performance test results: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 113/290
LTE Optimization Handbook TLA6.0
GridSum/Log
ServingR SRP
ServingS INR
DlPDCPTh roughput
MGR/TIPS/NEA
UlPDCPThr oughput
SFChangshaR oad1-NC-61.167 30.7581 92091.9 2088.51 DL_wo MG SFChangshaR oad1-NC-DL-63.214 30.8254 67797.3 1510.83 with MG period 40ms SFChangshaR oad1-NC-DL-65.538 32.7016 83996.2 1845.16 with MG period 80ms SFChangshaR oad1-NC-61.907 35.4302 148.295 7450.19 UL_wo MG SFChangshaR oad1-NC-UL-66.597 40.7813 46.6655 1973.73 with MG period 40ms SFChangshaR oad1-NC-UL-66.181 39.0265 94.6265 4630.84 with MG period 80ms Table 10-4: MG impact on DL&UL throughput performance of commercial deployment Qingdao - Urban)
NbrDlGrant PerSecond
NbrUlGrant PerSecond
Degrade rate
744.079
158.474
-
526.729
50
26.38%
654.92
109.679
8.79%
258.006
183.358
-
78.8394
50
73.51%
167.485
117.512
37.84%
UL/DL config2/7 (CMCC TDD LTE pre-
Based on the test result, we can find that the UE throughput degrade degree of 40ms MG pattern is larger than that of 80ms MG pattern for both DL and UL performance. And we can also find that MG does more aggressive negative impact on TDD UE UL throughput than its’ DL throughput performance. Index 1 2 3 4 5 6 Table 10-5: HO delay test Qingdao - Urban)
Test w/wo MG Handover Delay(s) InterFreq HO_with 0.02 80ms MG InterFreq HO_with 0.024 80ms MG InterFreq HO_with 0.025 40ms MG InterFreq HO_with 0.025 40ms MG 37900 IntraFreq HO 0.017 37900 IntraFreq HO 0.022 results w/wo MG enabled (CMCC TDD LTE pre-commercial deployment
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 114/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Based on the test result, we can find that, there is no visible gap on HO delay between 40ms MG and 80ms MG, although the HO delay of inter-freq HO may be longer than intra-freq HO case. Actually, using inter-freq will help to decrease interference of neighbour cells which leads to improvement of SINR, thus can help to improve both DL&UL throughput performance. Whether choosing inter-freq while optimizing throughput performance, we shall consider the total impact of using inter-freq, i.e. the negative impact of measurement gap, the improvement of SINR and the impact of HO performance.
10.1.15.1.1
CONFIGURATION USED
eNB Software version
TLA 6.0 ENB_TA0600_D20_E00029+ database MIM 15.1.3
UE
Hisilicon-E5776
CDS CDS7.0_35B + Hisi UE Agent v1.0.26.0 Table 10-6: Test SW configuration Reference
10.2 FEATURE LINKED 10.2.1 T115742- REL9 DL MU-MIMO BF (TM8 RANK1) 10.2.1.1
HIGH LEVEL DESCRIPTION AND BENEFITS:
This feature enables the eNB to transmit data to different UE on the same time-frequency resource, which is an advance technology o further improve the spectrum efficiency of LTE system. Base Station
Rx1 Tx1
X1
F2
Decoder W1
Tx2
Y1
Rx2 X2
Tx3 F1 Tx4
Decoder W2
Y2
Figure 10.2-1: MU-MIMO Principle The feature T115742 is targeted for customer CMCC. The working assumption is to use the installed 4+4 dual-polarized antenna array of TD-SCDMA with TD-LTE. The concept is to avoid complex signal processing in downlink transmission while basing the existing commEBB and EBB beamforming single-user beamforming algorithm. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 115/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
eNB intelligently pair two UE with distant DoA on the same frequency/time resource. Each multiplexed UE still use single user beamforming weightings with commEBB and EBB to maximize the signal strength and suppressing the inter-user interference. Adaptive MIMO Multiplexing do fast-switch SU-MIMO and MU-MIMO is supported, here lists 3 typical scenarios which Adaptive MU/SU-MIMO is operating: Scenario 1: Cell-Center ( High SINR ) UEs with MU-MIMO pairs -> MU-MIMO Scenario 2: Cell-Center ( High SINR) UEs w/o MU-MIMO Pairs -> SU dual layer BF Scenario 3: Cell-Edge ( Low SINR ) Ues -> SU single layer BF In all these three scenarios, two UEs are connected to the same eNB are placed at different DoA in respect to the antenna array.
10.2.1.2
HOW TO ACTIVATE:
MU Mode is introduced to TM8 (TDD only), so a new MIM parameter tm8MimoMultiplexingMode is introduced. When tm8MimoMultiplexingMode is set to “DualLayer SU-MIMO”, all multi user behaviour in the system shall be disabled, System falls back to L114531. Object
Attribute
UEAdaptiveBeamFormingTM8
tm8MimoMultiplexingMode
LteCellTDD transmissionMode Table 10-7: Parameters to activate feature
10.2.1.3
Value DualLayerMU-MIMO (or DualLayerSUMUAdaptive) TM8
FEATURE IMPACTS ON KPI’S & TUNABLE PARAMETERS:
Feature Impacts on KPI’s Mainly this feature will impact Cell DL total throughput due to the resources allocation multiplexed on the same time/frequency resource, which may help to improve the spectrum efficiency. Note that, control region cannot be multiplexed, so this feature does not affect the system capacity. Tuneable Parameters Ideally some exercise should be defined for tuning some of the parameters… in this case; (default values presented below). The tuning purpose is to maximize the throughput of both 2 paired UEs and also the cell total throughput.
Object
Attribute
Value Range
UEAdaptiveBeamFormingTM8
maxDoAGapThresholdMuPairing maxMuPairsPerUe minDoAGapThresholdMuPairing minMuPairsPerUe minSinrThresholdforMuMIMO retxTBMuEnable
[0,…,180] [1,…,20] [0,…,180] [1,…,20] [-10.0,…,30] false, true
Default Value 100 9 30 1 4.0 true
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 116/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
sinrCompensationforMuMIMO
[-10.0,…,30]
-4.0
Table 10-8: Tuneable parameters
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 117/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
11 UPLINK THROUGHPUT OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is Uplink Throughput in LTE. Normally some questions arise, such as: When to perform UL Throughput optimization? What method to apply? Equipments and tools to use? Which parameters can help improving UL Throughput?
Mainly the UL t-put optimization can occur when the average throughput value for a specific location is not matching the ALU product specification for a determined Bandwidth target. Before starting playing with the parameterization; usually is part of best practice rules for in Near Cell /Mid Cell & Cell Edge test to follow up simple steps as: Check the CQI Check the MCS triggered Check the UL BLER Evaluate RSSI vs. SNR relation Evaluate RSRP vs. RSRQ
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If regardless of the correct values, still facing some performance issues, they below parameters can be used in order to correct the situation. When changing parameters; you can adopt a more error-free approach, meaning that a parameter is changed at each time. If three or four parameters are changed same time… it could be difficult to understand which one is bringing the improvement in performance. As note; please remember that this can be a static test in each position, or can be a moving test… the same principles can be applied in both situations.
11.1 PARAMETERS OPTIMIZATION FOR IMPROVING UPLINK THROUGHPUT 11.1.1 UPLINKSIRTARGETVALUEFORDYNAMICPUSCHSCHEDULING This parameter used inside PUSCH power control algorithm as outer loop power control. It is initial target for the SINR values. During transmission, the SINR targets are changing based on the measured PL. The input of the UL outer-loop power control function is the PL along with some other parameters. In trial mode, the old value of this parameter is 10, but higher values, as high as 16, could be used for maximum throughput tests, especially outdoor for forcing higher power and thus allowing the compensation of higher PL for preserving the throughput. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 118/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
It is not recommended to consider high values in a loaded network due to increased interference even if the throughput for one user could be higher for higher values of this parameter. In high path loss conditions, the throughput with a lower SIR target becomes better because eNB will grant a lower MCS but with more PRBs than for high SIR Target. Default Value = “22” if Fractional Power control alpha factoris set to 0.8.
NEA Recommended Value= “It depends” if Fractional Power control is used or not, refer to Table 11-1.
PUSCHPowerControlAlphaFactor uplinkSIRtargetValueForDynamicPUSCHscheduling 1.0 22.0 0.8 16.0 0.7 14.0 Table 11-1: uplinkSIRtargetValueForDynamicPUSCHscheduling vs. PUSCHPowerControlAlphaFactor FFS: The field test for p0NominalPUSCH tuning still has not been performed yet, the results will be provided after the test cases are implemented. Expected behaviour when changing this parameter Increasing the value of this parameter would: Increase the transmission power of the UE which would result in using higher MCSs and obtaining higher throughputs. If the default setting of pUSCHPowerControlAlphaFactor is used, then the SIR target will be the same irrespective of the path loss and the UE power will be kept high. The UE power will reach its maximum value for lower path losses and thus the life of UE battery will be decreased. Reach the highest UE power for a lower pathloss which, in extremis, would limit the coverage. Increase interference in the neighbour cells due to higher transmitting power. This would result in lower throughputs in the neighbour cells. Decreasing the value of this parameter would: Decrease the power of the UE which would result in lower MCSs and lower throughputs. Decreasing this parameter will decrease the overall level of interference and hence improve the throughput of cell-edge users at the expense of cell-centre UEs, i.e. the peak throughputs will be lower. KPI Impact: Throughput: high values increase the throughput for near cell and mid-cell conditions. Capacity - high values allow reaching the capacity for a wider range of propagation conditions.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 119/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
For optimizing the value of this parameter for minimizing the interference in neighbour cells while maximizing the throughput in the analyzed cell, two cells and several UEs are needed (in the interfering cell). The following steps must be performed: Step 1: In victim cell, use the default parameters. In the interfering cell, set uplinkSIRtargetValueForDynamicPUSCHscheduling to one of the values {16, 14, 12, 10, 8}. Step 2: In victim cell perform an UL data transfer and log data related to PC command (TPC command field or F value) and PUSCH BLER and PUSCH DM RS SINR. Step 3: In the interfering cell, with UEs located near the cell edge, perform UL data transfer with several UEs, in synchronous manner and log the value of the throughput. Step 4: In the interfering cell choose another value for uplinkSIRtargetValueForDynamicPUSCHscheduling and repeat Step 2 and Step 3. Step 5: Post process the data and choose the value of uplinkSIRtargetValueForDynamicPUSCHscheduling that provides an acceptable trade-off between the throughput in the interfering cell and the throughput in the victim cell.
PUSCH TxPower (dBm)
uplinkSIRtargetValueForDynamicPUSCHsched uling impact PUSCH TxPower with Ncell no load 15 10
targetsinr8
5
targetsinr10
0
targetsinr12
-5
targetsinr14
-10 (-93,-89)
(-89,-85)
(-85,-81)
(-81,-77)
(-77,-73)
targetsinr16
RSRP (dBm)
Figure 11.1-1: Impact of PUSCH Power for ul_PUSCH_SIRtarget (Field Results CMCC LST-Urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 120/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
uplinkSIRtargetValueForDynamicPUSCHscheduling impact Ncell IoT with Ncell no load 14.00 12.00
IoT (dB)
10.00 IoT_SIR16
8.00
IoT_SIR14
6.00
IoT_SIR12
4.00
IoT_SIR10
2.00
IoT_SIR8
0.00 (-73,-77)
(-77,-81)
(-81,-85)
(-85,-89)
(-89,-93)
Serving Cell RSRP (dBm) Figure 11.1-2: Impact of the Neighbor Cell IoT for ul_PUSCH_SIRtarget (Field Results CMCC LSTUrban) Note: There are lots of result samples, so we calculated the average values according to RSRP range, that's why the RSRP is in pairs. Based on the above results in Figure 11.1-1 we can conclude that the higher the PUSCH SIR target the higher the UE Txpower at the same pathloss condition (the same RSRP range) to try reaching the target SIR. Based on the above results in Figure 11.1-2 we can conclude that the higher the PUSCH SIR target the higher the interference to the neighbour cells at the same pathloss condition (the same RSRP range) because UE Txpower is higher and cause the interference increased. These test results and conclusions are what we expected according to UL power control theoretical analysis. Note: Because of the tightly connected with other FPC parameters: pUSCHPowerControlAlphaFactor, pathLossNominal, p0NominalPUSCH, minSIRtargetForFractionalPowerCtrl and maxSIRtargetForFractionalPowerCtrl, parameter uplinkSIRtargetValueForDynamicPUSCHscheduling should be tuned cross with these parameters to improve UL performance, i.e. the suitable consistent set of these parameters configures the right Fractional Power Control according to the scenario.
11.1.2 PUSCHPOWERCONTROLALPHAFACTOR Part of PUSCH power control and is intended to allow partial compensation of the path loss or otherwise stated it allows controlling, by decreasing, the PUSCH power for users in cell edge conditions. If this parameter is set to 1, an increase in path loss will determine the same increase in PUSCH power. If this parameter is not set to 1, the increase in PUSCH power can be lower than the Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 121/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
increase in path loss. It is thus a means of controlling the UL interference created in the neighbour cell by the UEs found near the cell edge. Because the value of this parameter represents a trade-off between minimizing interference and maximizing throughput, its value must be set according to the client’s desired network behaviour. It is a key RF parameter. If Fractional Power Control is used, the recommended value of this parameter should be [0.0, 1.0]. If Fractional Power Control is not to be used, the parameter must have the value 1.
Recommended & Default Value= "1.0" if Fractional Power control is not used.
NEA Recommended Value= “1.0” to reach peak UL performance in trial mode; but “0.0
The optimization process of this parameter should include the customer definition of the optimum trade-off between cell throughput and interference towards the neighbour cells. The choice can be different if cell wise optimization is to be performed or if a network wide setting is being aimed Step 1: In interfering cell, 3 UEs located respectively in Near cell/Middle cell/Cell edge radio conditions and perform UL full buffer data transfer in synchronous manner. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 122/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 2: In victim cell, set the pUSCHPowerControlAlphaFactor to 1 and connect the UE in Near-Cell radio conditions. For a closer view to the lab results IoT should be considered. Step 3: In victim cell, start a UL full buffer UDP transfer and start driving from Near-Cell towards Edge-Cell. For a more consistent data we recommend a drive back as well logged in another trace. Step 4: Using the same cell and same route choose another value for pUSCHPowerControlAlphaFactor = {0.9, 0.8, 0.7, 0.6, 0.5} and repeat Step 1 and Step 2. Step 5: Post process the logged data and provide results in terms of UE TxPower, UL Throughput, UL IoT and PUSCH BLER of both victim cell and interfering cell.
FFS: The test for pUSCHPowerControlAlphaFactor tuning (in field) has not been completed yet, the results will be provided after the test cases are implemented. Hereunder are results from FDD.
Figure 11.1-3: Impact of the pUSCHPowerControlAlphaFactor =1.0 in MCS usage.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 123/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 11.1-4: Impact of the pUSCHPowerControlAlphaFactor =0.8 in Throughput per RB
Figure 11.1-5: Impact of the pUSCHPowerControlAlphaFactor =0.7 in Throughput per RB.
11.1.2.1
PUSCHPOWERCONTROLALPHAFACTOR COMBINATION TESTS
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 124/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Our initial assumption was to demonstrate other parameter presents in the power control algorithm performance impact, besides pUSCHPowerControlAlphaFactor parameter. So far we have considered set of values for related test cases, they can be seen in the Table 11-2 Parameter uplinkSIRtargetValueForDynamicPUSCHscheduli ng pUSCHPowerControlAlphaFactor pUSCHPowerControlAlphaFactor pUSCHPowerControlAlphaFactor pUSCHPowerControlAlphaFactor pUSCHPowerControlAlphaFactor pathLossNominal minSIRtargetForFractionalPowerCtrl maxSIRtargetForFractionalPowerCtrl Table 11-2: Different Set’s Combinations
SET1
SET2
SET4
SET3
SET5
SET6
SET7
15
15
15
15
15
15
15
1 0,8 0,7 0,6 0,5 60 0 15
1 0,8 0,7 0,6 0,5 60 0 20
1 0,8 0,7 0,6 0,5 60 0 19
1 0,8 0,7 0,6 0,5 85 0 20
1 0,8 0,7 0,6 0,5 70 0 19
1 0,8 0,7 0,6 0,5 20 0 19
1 0,8 0,7 0,6 0,5 40 0 19
FFS: The test cases for pUSCHPowerControlAlphaFactor combination tuning (in Lab) have not been completed yet, the results will be provided after the test cases are finished. We have considered a low IoT value=3dB in an UL&DL balanced cabled tests lab environment. We increased the attenuation over the same amount of time for all the tests we have performed, 20MHz and 2.6GHz network. In the theory, FPC function is same for TDD and FDD, based on Network Engineering for Optimization team’s tests, the obtained below results were from FDD: Set 1 Configuration analysis
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 125/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 11.1-6: Set 1 Result Based on the above results in Figure 11.1-6 we can conclude that having a non fractional power control mechanism will lead to best throughput (compared to FPC – ON cases) in near & mid cell. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 126/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Starting path loss 110dB we have observed an UL throughput degradation which corresponds to number of PRB decrease, which explains the product reaction to UL TxPower saturation (number of PRB are decreased when the TxPower is maximum and the number of PRB-es is still high).
Figure 11.1-7: Set 1UL Throughput & UE TX Power vs. Path loss NOTE: PL de-synch(~4dB) is due to the fact that in Figure 11.1-7 displays UL PL and Figure 11.1-10 displays DL PL. Please use the UL throughput curves to align the TxPower it effect. Comparing the plots above we can conclude that for pUSCHPowerControlAlphaFactor=0.7 number of PRB is not degraded, UE TxPower at near&edge cell environment has an acceptable value and also the uplink throughput still is satisfactory for most of the services an operator might offer. Below in Figure 11.1-8 is the theoretical analysis for SIR target calculation in case of SET2&4. Results confirm the theory and the fact that having Max_SIR_targetForFPC greater than SIR_Target_Nominal will not help (if Path loss nominal is equal 60dB). There are no commercial networks in which the less than 60dB is found. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 127/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 11.1-8: SIR Target for theoretical assumptions with different alpha factor values Looking to theoretical assumptions we have found that SET3 of values as something that could produce a similar UL throughput in Near cell environment compared to non-FPC case. As well as PL increases the UL throughput will smoothly decrease (around 93dB UL PL, compared to 110dB SET1&Alphafactor=1)
Figure 11.1-9: UL SIR Target for theoretical assumptions with different alpha factor values
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 128/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 11.1-10: Different alpha factor comparison (Throughput, PRB’s, SINR & PUSCH SINR Target) Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 129/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
NOTE: Comparing the SIR target collected with SIR target estimated in the theoretical approach. This proves that Alcatel-Lucent power control mechanism works as expected. The below comparison shows the UL throughput and TxPower comparison between the AlphaFactor=1 and AlphaFactor=0.7 for SET3. Out of this we can conclude that a SET3 (AlphaFactor=0.7) could be used when customers are requesting an high UL throughput in Near cell with a UL throughput decrease in late-mid cell and edge cell.
Figure 11.1-11: UL Throughput & UE TX Power vs. Path loss alpha factor 0.7 & 1 with set 3 Notes: Alpha Factor=0.7 is a good compromise between UL Throughput and the TxPower used. Never the less, bigger the AlphaFactor, bigger the UL throughput for near&mid cells. Bigger the AlphaFactor, lower the UL throughput for edge cells (below ~110dB UL PL). Nominal path loss impacts the UL throughput performances. Bigger the value, later the UL throughput decrease effect, but high the TxPower. Having Max_SIR_targetForFPC greater than SIR_Target_Nominal will not help(if for e.g. Path loss nominal is equal 60dB, will not help to have max 20dB when target is 15dB). Path loss nominal with value less than 60dB, is not useful. There are no commercial networks in which the less than 60dB is found. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 130/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SET 1 &AlphaFactor=1 is the template. It is a good setting combination for non-FPC and best performances in Near&Mid-Cell radio condition. We recommend this to be used for best performances. SET1 & AlphaFactor=0.7 is a good trend between the UL throughput and UL interference. It is more suitable for commercials networks were customer vision needs a near&mid-cell not targeting the maximum or what AlcatelLucent product can offer to respect of lower interference. SET3 & AlphaFactor=0.7&PathLossNominal = 85 looks a better UL throughput approach having as well the FPC on. So high throughput in near&mid-cell radio conditions, but as well higher throughput in edge cell (compared to nonfractional power control). TxPower is a compromise between the SET1&AlphaFactor=1 and SET1&AlphaFactor=0.7 in Figure 11.1-14: UE TX Power vs. Path loss for set1 & set3 with alpha factor 0.7 and 1.
Figure 11.1-12: UL Throughput & UE TX Power vs. Path loss for alpha factor 0.7 for all sets
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 131/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 11.1-13: UL Throughput vs. Path loss for set1 & set3 with alpha factor 0.7 and 1
Figure 11.1-14: UE TX Power vs. Path loss for set1 & set3 with alpha factor 0.7 and 1
11.1.2.2 PUSCHPOWERCONTROLALPHAFACTOR COMBINATION TESTS (LIVE NETWORK) FSS: The tuning test for pUSCHPowerControlAlphaFactor combination has not been implemented, the results will be provided after the test cases are performed.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 132/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
11.1.3 ULSCHEDPROPFAIRALPHAFACTOR ulSchedPropFairAlphaFactor: fairness factor of the UL scheduler ulSchedPropFairAlphaFactor = 1 yields a maximum C/I scheduler. The scheduler provides more resources to UEs in better conditions ulSchedPropFairAlphaFactor = 0.5 yields a fair scheduler. The scheduler attempts to provide equal share of RB utilization to all the UEs ulSchedPropFairAlphaFactor = 0.0 yields an increased fairness scheduler. The scheduler attempts to allocate the resources in such a way that all the UEs eventually get the same data rate
Default Value= "0.5" ulSchedPropFairAlphaFactor is tested with the following procedure: Step Step Step Step Step Step
1: ENB configuration:20M, sa1+ssp7, TM2, and uplinkControlChannelLUTindex=1; 2: The UEs’ distribution is depicted in Figure 10.1-43; 3: Set ulSchedPropFairAlphaFactor = 1; 4: All UEs are attached and perform fullbuffer uplink FTP transmission, the related data are recorded. 5: All UEs are attached and perform 3M (or 6M) uplink UDP transmission, the related data are recorded. 6: reconfigure alphaFairnessFactor with the value 1 and 2, repeat step 4&5 respectively.
In the Figure 11.1-15 it is presented the impact of using the different values for 11.1.3 ULSCHEDPROPFAIRALPHAFACTOR; such as 0 or 1.
Figure 11.1-15: Impact of the ulSchedPropFairAlphaFactor The test represented in the Figure 11.1-15 was made with the typical configuration; 5NC; 5MC; 5CE, and having the different ulSchedPropfairAlphaFactor will make the scheduler to adjust more or less fair the resources to all the ue's... Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 133/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Throughput – High values increase the throughput in the near&mid-cell condition at expense of cell edge users. Capacity - High values increase the cell overall throughput at expense of cell edge users..
11.1.4 MEASUREMENT GAP Please refer to section 10.1.15.
11.1.4.1.1 CONFIGURATION USED Please refer to section 10.1.15.1.1.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 134/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12 LATENCY OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is Latency in LTE. Normally some questions arise, such as: When to perform Latency optimization? What method to apply? Which parameters can help improving Latency?
Mainly the Latency optimization can occur when the average value for a specific location is not matching the ALU product specification in terms of KPI. Before starting playing with the parameterization; usually is part of best practice rules for in Near Cell /Mid Cell & Cell Edge test to follow up simple steps as: Check the CQI Check the MCS triggered Check the UL BLER Evaluate RSSI vs. SNR relation Evaluate RSRP vs. RSRQ
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If regardless of the correct values, still facing some performance issues, they below parameters can be used in order to correct the situation. When changing parameters; you can adopt a more error-free approach, meaning that a parameter is changed at each time. If three or four parameters are changed same time… it could be difficult to understand which one is bringing the improvement in performance. As note; please remember that this can be a static test in each position, or can be a moving test… the same principles can be applied in both situations.
12.1 PARAMETERS OPTIMIZATION LATENCY Latency is generally considered either as control plane latency or as user plane latency. Control plane latency involves the network attachment operation while user plane latency only considers the latency of packets while UE is in connected state Different kinds of latency definition are categorized as follows: Attach Latency (IMSI and GUTI) Control Plane Latency (idle-to-active) Detach Latency HO Gap (to be discussed during HO) User Plane Latency (ping RTT)
This chapter contains optimization guidelines for some parameters that have the highest impact on latency. LTE latency test cases include air latency and end-to-end latency as shown in Figure below. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 135/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
RAN / air latency in this document is defined in terms of the round trip time between UE and eNB.End to end latency is the elapsed round trip time from UE to a certain application server in the network. The EPC latency is the round trip time between UE and PDN GW.
Latency should be measured for UE with pre-scheduled and non-pre-scheduled. The E2E latency is depending on the location of the application server for the test, e.g. is it in internet world or is it a server co-located to the eUTRAN or possibly even a virtual server located inside the MME. Control Plane Latency The Idle to Active Transition Time is defined as the time required for the UE to transition from RRC idle state to RRC Active state. This is according to the 3GPP requirements stated: “Transition time (excluding downlink paging delay and NAS signalling delay) of less than 100 ms from a camped-state i.e. Idle Mode, to an active state, in such a way that the user plane is established” Additionally the latency between RA Preamble (followed by RRC Connection Request message NAS Service Request as shown in figure below) and RRC Connection Reconfiguration Complete message (Control Plane Latency) User plane latencies Ping is being used for discovering the User plane latency. Typically 32, 1000 and 1500 bytes payload sizes are set for the unloaded scenario, 32 and 1500 bytes for the loaded scenario Result will be the round trip delay between the eNodeB and the APP server An estimation of the RTD between the eNodeB and the APP server was performed by pinging the APP.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 136/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1 TEST RECOMMENDATION AND RESULTS 12.1.1.1
ATTACH LATENCY
For a successful attach attempt to happen, the needed the RACH preamble must be correctly decoded by eNB. Setting preambleInitialReceivedTargetPower >= dBm-104 will speed attach time but may result in interference to the other cells. Also making sure the following parameter preambleTransmitPowerStepSize= dB6 is important for reducing number of RACH preambles sent for RACH msg 1, and implicitly the attach time. The attach time latency is mainly dependent on the authentication procedure which has the highest proportion of total attach time. The IMSI attach time as an example is 600ms out of a total time of 677ms.
Figure 12.1-1: Example for IMSI attach procedure (with authentication)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 137/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-2: Example for GUTI attach procedure (no authentication) The examples are suggestive and clear that the RRC UE Capability procedure is time consuming. In conclusion attach time latency is mainly dependent on the authentication time.
12.1.1.2
ATTACH LATENCY RESULTS FROM CMCC LST
12.1.1.2.1 ATTACH LATENCY RESULT IN DIFFERENCT SCENARIOS The table below one can have an idea about the initial Attach time latency. The first row of the table indicates the value setting (disable/enable) of “isIntraFreqMobilityAllowed”, the second is the setting (5ms/20ms) of “aUGtriggerDelayforRACHmsg4”, the third is the setting (disable/enable) of NAS “Identity” procedure, the last one is the setting (disable/enable) of NAS “Authentication” procedure. When Authentication and Identity procedure are enable, attach latency is 476ms; When Authentication and Identity procedure are disable, attach latency is 280ms; If parameter isIntraFreqMobilityAllowed and aUGtriggerDelayforRACHmsg4 are optimized, the attach latency will be decreased to 260ms.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 138/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-3: Initial Attach latency comparison overview Attach latency with Authentication and Identity According to MIM default setting, if NAS Authentication and Identity are enabled, the Attach latency is 476ms,the proportion of authentication and identification procedure is 69%.
Figure 12.1-4: Initial Attach latency and ratio (with authentication and Identity) in CMCC LST Attach latency without Authentication and Identity According to MIM default setting, if NAS Authentication and Identity are disabled, the Attach latency is 280ms,the proportion of UE capability is 65%.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 139/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-5: Initial attach latency example (without authentication and Identity) in CMCC LST Attach procedure and latency
Figure 12.1-6: Initial Attach procedure and latency example (with authentication and Identity)
12.1.1.2.2 PARAMETER AUGTRIGGERDELAYFORRACHMSG4 IMPACTS ON LATENCY
Figure 12.1-7: aUGtriggerDelayforRACHmsg4 Vs. Attach latency Latency varies as the value of parameter aUGtriggerDelayforRACHmsg4 setting, the larger the value is set, the longer the latency. Based on the test results, we found that when the value of aUGtriggerDelayforRACHmsg4 is set from 20ms to 5ms, the latency will be decreased about 10ms.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 140/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.2.3 PARAMETER ISINTRAFREQMOBILITYALLOWED IMPACTS ON LATENCY
Figure 12.1-8: isIntraFreqMobilityAllowed Vs. Attach latency Based on the test results, we found that when the parameter isIntraFreqMobilityAllowed is set from enable to disable, the latency will be decreased about 10ms.
12.1.1.2.4 AUTHENTICATION CONFIGURATION IN MME In MME, authentication policy can be set separately for 6 service types as following: IRAT TAU: Inter RAT Track Area Update SERVCE REQUEST:UE initials service SUB ATTACH:UE Attach TAU update:Track Area Update UEint Detach:UE detach UE intEXTsevreq Parameter AUTHENTICATION INTERACTION defines the percent of authentication, for example: AUTHENTICATION INTERACTION 0%: means no authentication AUTHENTICATION INTERACTION 50%: means 50% procedure should adapt authentication AUTHENTICATION INTERACTION 100%: means always adapt authentication
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 141/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-9: Authentication configuration in MME
12.1.1.3 QUICK REMINDER FOR ISINTRAFREQMOBILITYALLOWED AND AUGTRIGGERDELAYFORRACHMSG4 12.1.1.3.1 ISINTRAFREQMOBILITYALLOWED Definition: This flag enables or disables the procedure of intra-frequency mobility. If disabled, the eNB will not trigger any outgoing intra-frequency mobility procedure and reject any incoming mobility procedure. Description: If the intra-frequency mobility is enabled via MIM configuration (i.e. isIntraFreqMobilityAllowed set to TRUE in Enb/ActivationService), the eNB will initiate a RRC Connection reconfiguration procedure after the RRC Connection establishment completion to setup measconfig for intrafrequency measurement.
12.1.1.3.2 ANTICIPATED UPLINK GRANTS (AUG) INTRODUCTION 12.1.1.3.2.1
PURPOSES FOR AUG: REDUCE LATENCY
Since TLA3.0/LA3.0, the ULS may receive requests for Anticipated Uplink Grants (AUG) from the Downlink Scheduler for two purposes: • In the context of a “RACH msg4/msg5 for Call Setup” scenario, trigger the transition of the UE context into synchronized state without waiting for the detection of the first UL SRS transmission. • Generate some UL grants without waiting for the successful completion of a “scheduling request UL grant for Buffer Status ReportBSR reportUL grant for awaiting data” scenario. AUG is not used in TLA3.0/LA3.0 outside the context of call setup scenarios (i.e. not used for bearer setup or handover cases).
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 142/290
LTE Optimization Handbook TLA6.0
12.1.1.4
MGR/TIPS/NEA
AUG PROCEDURE
AUG request transmission is triggered by the following events in the DL Scheduler: 1. Detection of a PUCCH ACK for the DL HARQ process associated to the transmission of a RACH msg4 at call setup. In this particular case: The AUG request from the DLS is either send by the DLS with a delay or processed by the ULS with a delay equal to AUGtriggerDelayforRACHmsg4 in order to avoid sending an UL grant before the UE has finished processing the RRC Connection Setup message. The AUG request is also used to force the UL UE context into synchronized state without waiting for the detection of the first UL SRS transmission. Detection of a PUCCH ACK for a DL HARQ process associated to the transmission of the last RLC PDU of a RRC message on SRB1 within a time window of duration AUGprocessDuration starting at the trigger of the first AUG request for the UE context.
2.
Upon receipt of an AUG request from the DL scheduler, the UL BO estimation function in the UL scheduler implements a series of UL BO increase on the logical channel associated to SRB1 with the following specificity: First UL BO increase: occurs upon receipt of the AUG request from the DLS Subsequent UL BO increases are every AUGulBOincreasePeriod ms until the max number of UL BO increase AUGulBOincreaseRepetitionNumber has been reached. The quantity of each of the UL BO increase on SRB1 is equal to: AUGulBOincreaseRAmsg4 for the first AUG request received from the DLS (i.e. the one triggered by the acknowledgement of the RACH msg4 on PUCCH). AUGulBOincreaseSRB1uponCallSetup for all the subsequent AUG requests received from the DLS.
Figure 12.1-10: AUG flow chart
12.1.1.5
AUGTRIGGERDELAYFORRACHMSG4
Description: Parameter aUGtriggerDelayforRACHmsg4 defines the AUG trigger delay implemented in the modem after detection of the RACH msg4. The delay is observed to avoid sending an UL grant before the UE has finished processing the RRC Connection Setup message (as it is unclear whether the UE monitors the PDCCH during this period).
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 143/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The parameter “aUGtriggerDelayforRACHmsg4" is recommaned to 20ms in default template. It could be adjust in kpi case "ueActivationTime", e.g. change to 10ms or even 5ms to shorten the delay between Msg4 & Msg5 for optimal performance.
12.1.1.6
IDLE TO ACTIVE LATENCY
The Idle to active latency is defined as the time needed for UE to enter active state. Starting point is the moment when UE initiates RACH Procedure with RACH message 1 for entering active mode until the UE sends RRC Connection Reconfiguration Complete message. For Idle to active the RACH procedure is included for this test. NOTE! The Idle to active transition is a process that needs less time than the normal attach procedure because the UE doesn’t need to initiate the authentication procedure anymore. If the RACH Preamble is not decoded by eNB then the UE will resend up to maximum preambleTransMax preambles before restarting RACH Procedure. The parameters: preambleInitialReceivedTargetPower, preambleTransmitPowerStepSize both should be set according to cell conditions for fast RACH procedure and hence a good Idle to active latency. According to the 3GPP requirements stated: “Transition time (excluding downlink paging delay and NAS signalling delay) of less than 100 ms from a camped-state i.e. Idle Mode, to an active state, in such a way that the user plane is established” Additionally the latency between RA Preamble (followed by RRC Connection Request message NAS Service Request as shown in figure below) and RRC Connection Reconfiguration Complete message (Control Plane Latency).
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 144/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-11: Idle to active message chart
12.1.1.7
IDLE TO ACTIVE LATENCY RESULTS FROM LAB TEST
The chart below one can have an idea about the total idle to active time latency from lab test,the average latency is 97ms,maximum is 111.7ms,minimum is 84.7ms.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 145/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-12: Example of total Idle to active latency
12.1.1.8
IDLE TO ACTIVE LATENCY RESULTS FROM CMCC LST
12.1.1.8.1 IDLE TO ACTIVE LATENCY RESULT IN DIFFERENCT SCENARIOS Idle to active latency result is according to parameter aUGtriggerDelayforRACHmsg4 (value: 20ms/5ms/4ms), parameter isIntraFreqMobilityAllowed(VALUE: enable/disable) and SINR (perfect>25dB,good>15dB,bad<5dB), the chart bellow shows results of each 18 scenarios.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 146/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 12.1-13: Idle to active latency results from CMCC LST The minimum latency is 85.8ms, it is happened when aUGtriggerDelayforRACHmsg4=4ms, isIntraFreqMobilityAllowed=disable, and SINR>25dB. The maximum latency is 107.1ms, it is happened when aUGtriggerDelayforRACHmsg4=20ms, isIntraFreqMobilityAllowed=enable.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 147/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.8.2 PARAMETER AUGTRIGGERDELAYFORRACHMSG4 VS. LATENCY
Figure 12.1-14: aUGtriggerDelayforRACHmsg4 vs. latency Latency varies as the value of parameter aUGtriggerDelayforRACHmsg4, the larger value set, the longer latency. When the value of aUGtriggerDelayforRACHmsg4 set from 20ms to 5ms, the latency will decrease about 10ms.
12.1.1.8.3 PARAMETER ISINTRAFREQMOBILITYALLOWED VS. LATENCY
Figure 12.1-15: isIntraFreqMobilityAllowed vs. latency When the parameter isIntraFreqMobilityAllowed set from enable to disable, the latency will decrease about 10ms. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 148/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.8.4 SINR VS. LATENCY
Figure 12.1-16: SINR vs. latency The latency is almost not affected by RF condition (SINR, RSRP, RI, etc).
12.1.1.8.5 LATENCY STEP BY STEP
Figure 12.1-17: Idle to activity message chart and latency Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 149/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.8.6 AUTHENTICATION CONFIGURATION IN MME In idle to activity procedure, authentication procedure may be included. Whether exist authenticaton is due to MME configuration. In MME, authentication policy can be set separately for 6 service types as following: IRAT TAU: Inter RAT Track area update SERVCE REQUEST:UE initials service SUB ATTACH:UE Attach TAU update:Track area update UEint Detach:UE detach UE intEXTsevreq If in SERVCE REQUEST item, Parameter AUTHENTICATION INTERACTION not set equal to 0%, then authentication should be adapt. Parameter AUTHENTICATION INTERACTION defines the percent of authentication, for example: AUTHENTICATION INTERACTION 0%: means no authentication AUTHENTICATION INTERACTION 50%: means 50% procedure should adapt authentication AUTHENTICATION INTERACTION 100%: means always adapt authentication
Figure 12.1-18: Authentication configuration in MME
12.1.1.9 QUICK REMINDER FOR ISINTRAFREQMOBILITYALLOWED AND AUGTRIGGERDELAYFORRACHMSG4 For detail information, please refer to section 12.1.1.3.
12.1.1.10 REFERENCE OF ALL THE SOFTWARE USED IN THE PLATFORM CMCC LST SW versions Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 150/290
LTE Optimization Handbook TLA6.0
HSS
R3.0 IAP1
PCRF-1
DSC_4_0_R3
S-GW
R3.1_R3
P-GW
R3.1_R3
MME02
R25.55.20
MGR/TIPS/NEA
UE
ENB_TA0600_D10_E00040 + database MIM 15.1.3 ed10 + NEM.LA6.0.1_D1.9 + WPS_MIM_11.4.1_ed05 Hisilicon-E398S, sw V100R001C00B201SP000,1.26.0
CDS
V7.0_B34121202 + Hisi UE Agent v1.0.26.0
eNB
EPS Integrity Protection Algorithm EPS Encryption Algorithm Table 12-1: SW Reference
128-EIA0 (No Algorithm) 128-EEA0 (No ciphering)
MME02
12.1.1.11 DETACH LATENCY When initializing a detach procedure the UE sends Detach Request after which eNB responds with Detach Accept. The detach latency is mainly clear context dependent and parties communication latencies. The Detach Latency should be around 200ms.
Figure 12.1-19: No RF parameter optimization possible for detach latency!
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 151/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
12.1.1.12 PING LATENCY RESULTS FROM CMCC LST In the table below, one can find the preliminary results for latency tests with 32, 1000, 1400 Bytes ping in CMCC TDD LTE large scale trial, 2.6 GHz, BW = 20MHz, uplinkControlChannelLUTindex=1, CFI=3. The Hisilicon UE has been tested with and without prescheduled for U-plane latency feature. Hisilicon UE
avg. Ping Latency 32 bytes
avg. Ping Latency 1000 bytes
avg. Ping Latency 1400 bytes
Location
prescheduled
non-prescheduled
Near Cell
15.4 ms
30 ms
Mid Cell
15.6 ms
31.4 ms
Cell Edge
17.2 ms
32.4 ms
Near Cell
16.4 ms
40.4 ms
Mid Cell
16 ms
42.4
Cell Edge
19.4 ms
44.6 ms
Near Cell
17 ms
40.4 ms
Mid Cell
18 ms
42.8 ms
Cell Edge 20.6 ms 47 ms Table 12-2: Ping Latency for 32 Bytes with and without prescheduled for U-plane latency (CMCC LST-Density urban)
Userplane Latency_32Bytes Non-prescheduled vs. pre-scheduled 35 30 Latency (ms)
32.4
31.4
30 25 20 15
17.2
15.6
15.4 10 5 0 high SINR
medium SINR Non Pre-scheduled
low SINR
Pre-scheduled
Figure 12.1-20: Non pre-scheduled and pre-scheduled Ping latencies with 32 byte payload (CMCC LST-Urban) Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 152/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The average latency increase with degraded SNR due to fading conditions, the range of variation is usually few ms. Pre-scheduled ping latency outperforms non pre-scheduled ones as expected, for all SNR range, in the 32 bytes case the range of this variation is around 15ms.
Userplane Latency_1400Bytes Non-prescheduled vs. pre-scheduled 50 45
47
Latency (ms)
40 35
42.8
40.4
30 25 20 15
20.6
18
17
10 5 0 high SINR
medium SINR Non Pre-scheduled
low SINR
Pre-scheduled
Figure 12.1-21: Non-scheduled and pre-scheduled Ping latencies with 1400 byte payload (CMCC LST-Urban) Obviously the expected behaviour of increased ping payload is increased latency which is reflected by comparing the 2 above bar charts.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 153/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
50
47
45 40 Latency (ms)
44.6
42.4
40.4
35 30
42.8
40.4
32.4
31.4
30
25 20 16.4
15
17
16
18
15.6
20.6 19.4 17.2
15.4
10 5 0 high SINR
medium SINR
low SINR
high SINR
Non Pre-scheduled 30 Bytes
medium SINR
low SINR
Pre-scheduled 1000 Bytes
1400 Bytes
Figure 12.1-22: Overall Results (CMCC LST-Urban)
Below the results for the Pre-scheduled and Non Scheduled ping latencies in High and Low SNR ranges for a more differential view of the latencies in case of UE situated in several cell coverage positions.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 154/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Pre-scheduled vs. Non-prescheduled Ping Latency 32 Bytes Low SNR vs. High SNR 60 49
Latency (ms)
50 40 30
35 30
35 30
35 30 25
21 20
48
45
31 27
26
33
31
32
25 19
18 18 16 14 14 15 14 15
16
32
31
18 14
14
17
18 14
SNR high
prescheduled 3
prescheduled 2
prescheduled 1
nonprescheduled 3
nonprescheduled 2
nonprescheduled 1
prescheduled 3
prescheduled 2
prescheduled 1
nonprescheduled 3
nonprescheduled 2
0
nonprescheduled 1
10
SNR low
Min
Max
Mean
Figure 12.1-23: Pre-scheduled vs. Non Scheduled Ping latency (32 Bytes) in Low SNR and High SNR (CMCC LST-Urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 155/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Pre-scheduled vs. Non-prescheduled Ping Latency 1400 Bytes Low SNR vs. High SNR 90
81
80
71
71
Latency (ms)
70 60 50 40
49
46 45 45 41 40 40 35 32 32
36
32
28
26
30
48
45
34
30
19 15 17 15 17 15 17
20
15
18
38 31 15
20
22 15
SNR high
prescheduled 3
prescheduled 2
prescheduled 1
nonprescheduled 3
nonprescheduled 2
nonprescheduled 1
prescheduled 3
prescheduled 2
prescheduled 1
nonprescheduled 3
nonprescheduled 2
0
nonprescheduled 1
10
SNR low
Min
Max
Mean
Figure 12.1-24: Pre-scheduled vs. Non Scheduled Ping latency (1432 Bytes) in Low SNR and High SNR (CMCC LST-Urban)
12.1.1.12.1
QUICK NOTES ABOUT THE TESTING
Prescheduled mode = disabled Non-Prescheduled mode = enabled, it is default set to enabled.
12.1.1.12.2 eNB Software version UE
CONFIGURATION USED TLA 4.0 ENB_TA0400_D00_E00027+ database MIM 11.4.1 ed05 + NEM.LA4.0.1_D1.9 + WPS_MIM_11.4.1_ed05 Hisilicon-E398S, sw V100R001C00B201SP000,1.26.0
CDS CDS7.0_15B120302 + Hisi UE Agent v1.0.26.0 Table 12-3: Test SW configuration Reference
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 156/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
13 CAPACITY OPTIMIZATION HINTS In this chapter it will be highlighted the main focus of testing and the primary steps that will allow to optimize a specific domain and the most important /priority parameters; in this case the domain addressed is Capacity in LTE. Normally some questions arise, such as: When to perform Capacity optimization? What method to apply? Equipments and tools to use? Which parameters can help improving Capacity?
By default, there is not much chance to make an improvement of the connected users… normally what can be worked is split of the performance (resource block allocation), for users in different radio conditions. This means that we can have fair or unfair resource distribution; for more details you can check the parameters described below.
13.1 PARAMETERS OPTIMIZATION FOR IMPROVING CAPACITY ENB capacity depends on the eNB configuration request, which is further in compliance with the customers’ request. Note that the capacity is also confined by the product itself with the capability restrictions of the hardware as well as the software inside. On the other hand, it has been theoretically proofed that there is tradeoff between wireless system throughput and capacity. Fairness is therefore introduced to represent how the system handles the scheduling of different users in various wireless channel conditions. Different fairness factors lead to different results with respect of capacity in the following section.
13.1.1 ENB CAPACITY CONFIGURATIONS Features and configuration elements impacting the eNB capacity figures are: TDD UL/DL Bandwidth configured through parameters bandwidth. Different capacity figures are available depending on the LTE bandwidth used. Parameter maxNbrOfUsers controls the number of users that can be admitted in a cell (users per modem). Parameter maxNumberOfCallPerEnodeB controls the number of users that can be admitted on the eNB (users per Controller board)
From TLA6.0, there is no different between 10/20 MHz bandwidth, different DL/UL configurations and 2 antennas/8 antennas, with parameter uplinkControlChannelLUTindex setting to value 0 and CFI setting to 3. The following table is summarizing the eNB capacity figures that can be obtained. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 157/290
LTE Optimization Handbook TLA6.0
TDD bandwidth n10020MHz/n5010MHz
MGR/TIPS/NEA
DL/UL Configuration 1/configuration2, 2/8 antennas, Special subframe 7/5 Max. Active user per cell = 100; Max. L2 user context per cell = 106; Max. data bearer per cell = 325; Average Scheduled #user per TTI per cell = 10
Table 13-1: TLA6.0 Capacity figures
13.1.2 ALPHAFAIRNESSFACTOR Parameter alphaFairnessFactor tunes the alpha fairness factor (thus the behaviour) of the DL scheduler. The scheduler is the processing entity that allocates resources to user plane and control plane traffic. Different factors such as channel quality, data pending in buffers, relative priorities in terms of QoS traffic are some key factors that are used by the scheduler to pick specific users from the active pool and allocate air-link resources to them. The overall goal is to ensure that users do get a chance to share the available bandwidth in the system, and the resources are allocated in an efficient manner while maintaining good system utilization.
Default Value= "1" Expected behaviour when changing this parameter alphaFairnessFactor = 0 Aggressive mode (α = 0) means that the scheduler prefers better radio channel conditions to normal radio channel conditions. Therefore, the scheduler provides more resources to UEs in better conditions. The better the radio conditions of the UE, the more resources (and hence the higher the data rate) it gets. Using test procedures described in section 10.1.9, the follow results when apply the aggressive mode you should expect: UE in Cell Edge are allocated the less PRB in average to favour UE with better radio conditions. On the other side, since the better radio condition usually brings throughput gains, this mode may have best throughput performance as compared to proportional fairness and enhanced proportional fairness. alphaFairnessFactor = 1 Proportional Fairness (α = 1), the scheduler attempts to provide the same number of RBs to all the UEs (despite their different conditions). Using test procedures described in section 10.1.9, the follow results when apply the proportionalfair mode you should expect: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 158/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Average fairness maintained despite different radio conditions alphaFairnessFactor = 2 Conservative mode (α = 2) shows the more Ue move away, the more PRB they are allocated, decreasing gaps between Ue Throughput but decreasing the cell Throughput as well. The scheduler attempts to allocate the resources in such a way that all the UEs eventually get the same data rate (which is not the case of the fair scheduler since different radio conditions result in different data rates even when the number of resources is the same, hence the increased fairness of the scheduler, as compared to the “regular” fair scheduler). Using test procedures described in section 10.1.9, the follow results when apply the conservative mode you should expect: When the UE’s is in bad radio conditions, more PRB are allocated in average to decrease throughput gaps Average UE DL throughput per radio group (Good, Medium, Bad radio conditions), decrease compared to proportionalfair mode In overall the final 15 UE throughput decrease compared with proportionalfair mode alphaFairnessFactor conclusions: Conservative mode (α = 2), shows the more UE move away, the more PRB they are allocated, decreasing gaps between UE Throughput but decreasing the cell Throughput as well. Aggressive mode (α = 0), shows the more UE come closer, the more PRB they are allocated, increasing gaps between UE Throughput but also the cell Throughput. Proportional Fair mode (α = 1), shows the better fairness distribution between the different radio conditions. All UE’s will have the same resources even in bad radio conditions, appropriate for commercial networks. KPI Impact: Throughput – Low values increase the throughput in the near&mid-cell condition at expense of cell edge users. Capacity - Low values increase the cell overall throughput at expense of cell edge users.
13.1.3 ULSCHEDPROPFAIRALPHAFACTOR Parameter ulSchedPropFairAlphaFactor tunes the alpha fairness factor (thus the behaviour) of the UL Scheduler. Default Value= "0.5" Expected behaviour when changing this parameter ulSchedPropFairAlphaFactor = 1 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 159/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Aggressive mode (α = 1), Yields a maximum C/I scheduler. The scheduler provides more resources to UEs in better conditions. The better the radio conditions of the UE, the more resources (and hence the higher the data rate) it gets. Using test procedures described in section 11.1.3, the follow results when apply the aggressive mode you should expect: ULS choice only depends on spectrum efficiency => unpredictable fairness behaviour TCP flow control increases BO of favoured Ues and decrease BO of others ulSchedPropFairAlphaFactor = 0.5 Proportional Fair (α = 0.5), Yields a fair scheduler. The scheduler attempts to provide the same number of RBs to all the UEs (despite their different conditions). Using test procedures described in section 11.1.3, the follow results when apply the proportional fair mode you should expect: PRB used in UL => decreased in degraded radio when Ue reach maximum path loss (PHR limitation) Fairness no more significant between Cell edge group and the others conditions, since UE is at maximum path loss. ulSchedPropFairAlphaFactor = 0 Conservative mode (α = 0), yields an increased fairness scheduler. The scheduler attempts to allocate the resources in such a way that all the UEs eventually get the same data rate (which is not the case of the fair scheduler since different radio conditions result in different data rates even when the number of resources is the same, hence the increased fairness of the scheduler, as compared to the “regular” fair scheduler). Using test procedures described in section 11.1.3, the follow results when apply the conservative mode you should expect: Average Ue Throughput per radio group => differences decreased compared to proportional Fair mode Cell Throughput decreased compared to proportional Fair mode The more UE are in bad radio, the more PRB they are allocated in average to decrease Throughput gaps. ulSchedPropFairAlphaFactor conclusions: Conservative mode (α = 0), does not show any improvement compared to proportional fair results, as shown in simulations results as well. Proportionalfair mode (α = 0.5), fairness no more significant between Cell edge and the other radio conditions, since UE is at maximum path loss. More PRB bandwidth used in this configuration compared with the other. Better configuration for commercial networks. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 160/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Aggressive mode (α = 1), is not a significant test since throughput between users are not predictable. This mode is not recommended for customer deployment. KPI Impact: Throughput: high values increase the throughput for near cell and mid-cell conditions. Capacity - high values allow reaching the capacity for a wider range of propagation conditions.
13.2 UL PHYSICAL CHANNELS CONFIGURATION AND CAPACITY ANALYSIS 13.2.1 SRS CONFIGURATION AND CAPACITY ANALYSIS 13.2.1.1
THE FUNCTION OF SOUNDING RS
Sounding Reference Signal (SRS) provides the uplink channel quality, which will be used to do fast link adaptation, at least for the following purposes: Interference management Enable frequency selective scheduling Selection of modulation and coding scheme Servers as a reference for closed loop Power Control Determine the downlink beamforming weights (for TDD TM7/TM8) Providing additional channel information to support channel dependent scheduling on the downlink Can be used for the judgment of the uplink synchronization
13.2.1.2
SRS CONFIGURATION IN CURRENT RELEASE
13.2.1.2.1 THE OBJECTIVES AND CONTENT OF SRS CONFIGURATION According to 3GPP TS36.331 section 6.3.2, the IE SoundingRS-UL-Config is used to specify the uplink Sounding RS configuration.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 161/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 13.2-1: SoundingRS-UL-Config information element Following Sounding Reference Signal (SRS) parameters are UE specific semi-statically configurable by higher layers: Transmission comb Starting physical resource block assignment Duration of SRS transmission: single or indefinite (until disabled) SRS configuration index ISRS for SRS periodicity and SRS subframe offset SRS bandwidth Frequency hopping bandwidth Cyclic shift Note that, a special subframe UpPTS equivalent to two normal sub-frames for SRS transmission, which means there could be 2 SRS symbols on UpPTS while only 1 SRS symbol on (per) normal UL sub-frame. In ALU current release, SRS is configured based on the following principle: SRS is always present. (from LA2.x) The SRS is transmitted on the last SC-FDMA symbol of the (UL) sub-frame (i.e. symbol location = 13). (from LA2.x) ackNackSRS-SimultaneousTransmission is TRUE. (from LA2.x) The SRS is allowed to overlap with the PUCCH AN/SR resource (if so, shortened PUCCH format 1/1a/1b is used). For example, 24 PRB SRS overlaps with 2PRB PUCCH configuration in the 5MHz BW. The SRS is allowed to overlap with PUCCH CQI resource. The SRS reading of the PUCCH CQI PRB is not useful for PUSCH transmission, but the SRS transmission from other users in the same TTI can cause a small interference (1 SC-OFDM symbol in the TTI) to P-CQI transmissions, which leads to a small degradation on the P-CQI performance. To overcome the impact, the P-CQI transmission power Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 162/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
can be boosted slightly by setting higher deltaFPUCCHFormat2x parameters. (from LA3.x) Only wideband SRS is configured by default. The supported SRS BW setting is per system bandwidth. However from TLA5.x narrowband SRS is also supported through activation flag dynamicSubbandSrsEnable. Cell-specific SRS subframe is configured on all (UL) subframes. (from LA3.x) Different SRS periods for different UEs are required per cell. (from LA3.x) Single CAZAC sequence is used per cell. (from LA2.x) Maximum 8 SRS sequences multiplexed per (UL) subframe. (from LA3.x) Both combs can be used per (UL) subframe with max 4 SRSs per comb. (from LA3.x) Default frequency position index is used for all UEs. From LA5.x srsFrequencyDomainPosition will determine the start position of the SRS when narrowband SRS use is enabled, else the existing hardcoded value of 0 will be used. (not present in TLA6.0 for the moment) Frequency hopping of SRS is not supported. (from LA2.x)
13.2.1.2.2 SRS BANDWIDTH CONFIGURATION AND SUGGESTION SRS bandwidth is configured by Cell-specific parameter srsBandwidthConfiguration (CSRS) and UEspecific parameter srsBandwidth ( b BSRS {0,1, 2,3} ). The SRS transmission bandwidths are a multiple of 4 RBs for all values and not include the PUCCH region. SRS allocation in frequency domain shows the frequency allocation method for SRSs with different bandwidths based on the proposed OVSF concept. Layer 1 1/2 BW 0 Layer 2
5MHz
5MHz
10MHz
Frequency
20MHz
1/4 BW 0 Layer 3 1/8 BW 0 1/8 BW 0
5MHz RS UEs
System bandwidth (BW 0)
5MHz RS UEs
10MHz RS UEs
20MHz RS UEs
Symbol
Figure 13.2-2: SRS frequency allocation example For TDD, the following configuration is used: If dynamicSubbandSRSenable is false, constant SRS bandwidth is assumed for different UL system bandwidth and OAM configuration is used: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 163/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
mSRS,b
= 96 for 20MHz system bandwidth. In case of PRACH format 4, the parameter srsMaxUpPts is set to enable, so the SRS on special subframe will be shorten to 80 PRBs;
mSRS,b
= 48 for 10MHz system bandwidth. In case of PRACH format 4, the parameter srsMaxUpPts is set to enable, so the SRS on special subframe will be shorten to 40 PRBs;
mSRS,b
= 72 for 15MHz system bandwidth. In case of PRACH format 4, the parameter srsMaxUpPts is set to enable, so the SRS on special subframe will be shorten to 64 PRBs. If dynamicSubbandSRSenable is true, all the possible srsBandwidthConfiguration and srsBandwidth (except 4 RB subband SRS) can be supported. mSRS,b
and
Nb
80 N UL 110
RB values for the uplink bandwidth of (20MHz) SRS-Bandwidth SRS-Bandwidth SRS-Bandwidth SRS-Bandwidth
SRS bandwidth configuration BSRS 0
BSRS 2
BSRS 1
BSRS 3
CSRS
mSRS,0
N0
mSRS,1
N1
mSRS,2
N2
mSRS,3
N3
0
96
1
48
2
24
2
4
6
mSRS,b
and
Nb
40 N
UL RB
60 (10MHz)
values for the uplink bandwidth of SRS-Bandwidth SRS-Bandwidth SRS-Bandwidth
SRS bandwidth configuration BSRS 0
BSRS 2
BSRS 1
SRS-Bandwidth
BSRS 3
CSRS
mSRS,0
N0
mSRS,1
N1
mSRS,2
N2
mSRS,3
N3
0
48
1
24
2
12
2
4
3
mSRS,b
and
Nb
60 N
80 (15MHz)
values for the uplink bandwidth of SRS-Bandwidth SRS-Bandwidth SRS-Bandwidth
SRS bandwidth configuration BSRS 0 CSRS
UL RB
mSRS,0
BSRS 2
BSRS 1
N0
mSRS,1
0 72 1 24 Table 13-2: TLA6.0 SRS bandwidth configuration
mSRS,2
N1 3
12
SRS-Bandwidth
BSRS 3
mSRS,3
N2 2
4
N3 3
NOTE: The number of supported sounding bandwidths is impacted by following issues: The power limitation on UE transmission The number of supportable sounding UEs The sounding bandwidth needed to get UL channel dependent scheduling gain Due to the path loss and power limit, the Cell edge users suggested to use the subband SRS, in contrast, the cell centre users suggested to use the wideband SRS. SRS bandwidth 96 80 72 64 60 48 40 19.823dB 19.031dB 18.573dB 18.062dB 17.782dB 16.812dB 16.021dB △PHR Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 164/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SRS bandwidth 32 24 20 16 12 15.051dB 13.802dB 13.010dB 12.041dB 10.792dB △PHR Table 13-3: Power limitation of different SRS bandwidth configuration
8 9.031dB
4 6.021dB
As there are different characters of wideband and subband SRS, the selection of souding scheme depends on balance of pros and cons. The simulation result will be good reference to determin souding scheme – both link level simulaiton and system level simulation.
13.2.1.2.3 SRS SUBFRAME & PERIODIC CONFIGURATION AND SUGGESTION SRS periodic is configured by cell specific subframe configuration and UE specific srs-ConfigIndex ISRS. The cell specific subframe configuration period and the cell specific subframe offset, relative to a frame for SRS in different UL/DL config case are listed in table below. From TLA3.0, the SRS subframe configuration is configured to 1 and 4 incase of UL/DL config 2 and 1, as highlighted in the table. srsSubframe Configuration Period Transmission offset Binary Configuration (subframes) (subframes) 0 0000 5 {1} 1 0001 5 {1, 2} 2 0010 5 {1, 3} 3 0011 5 {1, 4} 4 0100 5 {1, 2, 3} 5~15 Table 13-4: SRS subframe configuration
TLA3.0 requirement Not required UL/DL config 2 Not required Not required UL/DL config 1 Not required
SRS periodicity defines the SRS period in ms and SRS Subframe Offset provides the information about the SRS transmission time offset within the SRS period. The SRS Periodicity and the SRS subframe offset are configured per each UE and signaled to the UE by srs-ConfigIndex ISRS, defined in Table below: Configuration Index ISRS
SRS Periodicity TSRS (ms)
SRS Subframe Offset
Toffset
TLA3.0 requirement
0~9 2 Not required 10 – 14 5 ISRS – 10 Not required 15 – 24 10 ISRS – 15 required 25 – 44 20 ISRS – 25 required 45 – 84 40 ISRS – 45 required 85 – 164 80 ISRS – 85 required 165 – 324 160 ISRS – 165 required 325 – 644 320 ISRS – 325 required 645 – 1023 reserved reserved Not required Table 13-5: ISRS - UE Specific SRS Periodicity and Subframe Offset Configuration for TDD Since TLA3.0, eNB shall support different SRS period values for different UEs connected at the same time onto the same cell. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 165/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Note that, besides the requirement of capacity, the configuration of SRS periodicity should also consider UE speed, scheduling behaviour, etc. Here under are some suggestions: For FDS user, the scheduling may base on wideband CQI with long CQI update cycle, and then SRS transmission period of each UE can be configured as a large value, such as 160ms; For FSS user, the scheduling may base on subband CQI with short CQI update cycle, and the CQI update cycle determines the time of the SRS transmission of the UE has completed the full bandwidth. However since the Tx power of the UE is limited, it is not necessarily able to guarantee maximum SRS bandwidth at the same time transmitting the SRS sequence, so the UE SRS transmission periodic are usually only the current scenario updated CQI cycle fraction of even a few tenth.
13.2.1.3
ANALYSIS OF SRS CONFIGURATION IMPACT ON CAPACITY
Besides the above SRS configurations, the comb of SRS and cycle shit of SRS also impact the supported UEs. According to 3GPP spec, one comb maximum support 8 users, and the current ALU version only supports 4 users. The number of users that can transmission the ZC sequence in the SRS is: Capacity_SRS = Nub. of comb per sub-frame * No. of multiplexed UE per comb * No. of UL subframe in periodicity of SRS Assume that, Uplink-downlink configuration1 + CSRS=0 & BSRS=0 + srsSubframeConfiguration=4 + 45 ISRS 84 (i.e. 40ms periodicity), then we can get the capacity supported by SRS configuration: Capacity_SRS= 2(combs per sub-frame)×4 (UEs percomb)×(2+1+1)(SRS symbols per 5ms SRS period) ×40/5 (number of SRS period) (UL subframes)=2×4×32=256 UEs per SRS period. The calculation of the assumed SRS configuration could be present as below:
Figure 13.2-3: SRS capacity calculation example Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 166/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The number of UE per SRS can be increased by modifying the following parameters: a) Using the Subband SRS (but the PUSCH capacity may be decreased) b) Increasing the periodicity of SRS transmission, but this might affect our uplink throughput. c) Increasing the number of UE multiplexed per RB (more accurately transmission comb). However, it is up to the operator’s traffic model to decide how to play with these values and achieve the goals planned.
13.2.2 PUCCH CONFIGURATION AND CAPACITY ANALYSIS 13.2.2.1
THE FUNCTION OF PUCCH
PUCCH (Physical Uplink Control Channel) carries a set of information called "UCI (Uplink Control Information)". Depending on the configuration, it can carry only a few of the following information: SR: Scheduling Request CQI: Channel Quality Indication PMI: Pre-coding Matrix Indication RI: Rank Indication ACK/NACK: For Downlink Data Periodic and aPeriodic mode UE reports its UCI to the eNB, eNB allocation report opportunity and resource index for every UE. Depending on what kind of information the UCI in PUCCH carries, PUCCH is classified into various formats as follows: PUCCH Format UCI information Format 1 Scheduling Request (SR) Format 1a 1-bit HARQ ACK/NACK with/without SR 2-bit HARQ ACK/NACK with/without SR Format 1b (This is for MIMO, 1 bit for each transport block) Format 2 CQI (20 coded bits) Format 2 CQI and 1 or 2 bit HARQ ACK/NACK - 20 bits - Extended CP only Format 2a CQI and 1 bit HARQ ACK/NACK - (20 + 1 coded bits) - normal CP only Format 2b CQI and 2 bit HARQ ACK/NACK - (20 + 2 coded bits) - normal CP only Table 13-6: Supported PUCCH formats According to 3GPP TS 36.213 section 10.1, the following combinations of uplink control information on PUCCH are supported: Format 1a for 1-bit HARQ-ACK or in case of FDD for 1-bit HARQ-ACK with positive SR Format 1b for 2-bit HARQ-ACK or for 2-bit HARQ-ACK with positive SR Format 1b for HARQ-ACK with channel selection Format 1 for positive SR Format 2 for a CQI/PMI or a RI report when not multiplexed with HARQ-ACK Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 167/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Format 2a for a CQI/PMI or a RI report multiplexed with 1-bit HARQ-ACK for normal cyclic prefix Format 2b for a CQI/PMI or a RI report multiplexed with 2-bit HARQ-ACK for normal cyclic prefix Format 2 for a CQI/PMI or a RI report multiplexed with HARQ-ACK for extended cyclic prefix
13.2.2.2
PUCCH CONFIGURATION IN CURRENT RELEASE
13.2.2.2.1 THE OBJECTIVES AND CONTENT OF PUCCH CONFIGURATION According to 3GPP TS36.331, the IE PUCCH-Config, CQI-reporting and SchedulingRequestConfig are used to specify the PUCCH configuration. The PUCCH related RRC layer parameters are carried through the RRC connection reconfiguration signalling. PUCCH-Config information elements -- ASN1START PUCCH-ConfigCommon ::= deltaPUCCH-Shift nRB-CQI nCS-AN n1PUCCH-AN }
SEQUENCE { ENUMERATED {ds1, ds2, ds3}, INTEGER (0..98), INTEGER (0..7), INTEGER (0..2047)
PUCCH-ConfigDedicated ::= ackNackRepetition release setup repetitionFactor spare1}, n1PUCCH-AN-Rep } } tdd-AckNackFeedbackMode OPTIONAL -- Cond TDD }
SEQUENCE {
CHOICE{ NULL, SEQUENCE { ENUMERATED { n2, n4, n6, INTEGER (0..2047) ENUMERATED {bundling, multiplexing}
-- ASN1STOP IE PUCCH-ConfigCommon IE Parameter RAN1 Parameter deltaPUCCH-Shift
PUCCH shift
nRB-CQI
(2) N RB
nCS-AN n1Pucch-AN
N cs(1) (1) N PUCCH
NOTE Cyclic shift length Number of PRBs reserved for 2/2a/2b Number of cyclic shifts used for 1/1a/1b in a resource block with mix formats Number of 1/1a/1b indices reserved for SRI and SPS
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 168/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Table 13-7: PUCCH common parameters IE PUCCH-ConfigDedicated IE Parameter
RAN1 Parameter
NOTE
ackNackRepetition
-
Parameter indicates whether ACK/NACK repetition is configured, see TS 36.213 [10.1].
repetitionFactor
see TS 36.213 [10.1] where n2 corresponds to repetition factor 2, n4 to 4.
n1Pucch-AN-Rep
see TS 36.213 [10.1]. Number of 1/1a/1b indices reserved for ACK/NACK repetition.
tdd-AckNackFeedbackMode
Parameter indicates one of the two TDD ACK/NACK feedback modes used, see TS 36.213 [7.3]. bundling corresponds to use of ACK/NACK bundling whereas, multiplexing corresponds to ACK/NACK multiplexing. The same value applies to both ACK/NACK feedback modes on PUCCH as well as on PUSCH. For TDD configuration 5, E-UTRAN should always set this field to bundling.
-
Table 13-8: PUCCH dedicated parameters SchedulingRequestConfig information element -- ASN1START SchedulingRequestConfig ::= CHOICE { release setup sr-PUCCH-ResourceIndex sr-ConfigIndex dsr-TransMax
NULL, SEQUENCE { INTEGER (0..2047), INTEGER (0..155), ENUMERATED { n4, n8, n16, n32, n64, spare3,
spare2, spare1} } } -- ASN1STOP UE is configured using IE SchedulingRequest-Configuration: IE Parameter RAN1 Parameter NOTE sr-PUCCHResourceIndex
(1 ) nPUCCH, SRI
Number of resource indices reserved for SR transmissions.
sr-ConfigIndex
-
This parameter determines that the UE specific subframe configuration period and subframe offset.
dsr-TransMax
-
The DSR's largest number of transmissions
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 169/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Table 13-9: SR configuration parameters The SR Periodicity and the SR subframe offset are configured per each UE and signaled to the UE by SR Configuration Index ISR, defined in Table below: SR configuration Index ISR 0–4 5 – 14 15 – 34 35 – 74 75 – 154 155 Table 13-10: SR configuration Index
SR periodicity (ms)
SR subframe offset
5 10 20 40 80 OFF
ISR ISR -5 ISR -15 ISR -35 ISR -75 N/A
CQI-ReportConfig information elements -- ASN1START CQI-ReportConfig ::= cqi-ReportModeAperiodic
SEQUENCE { ENUMERATED { rm12, rm20, rm22, rm30,
rm31, OPTIONAL, nomPDSCH-RS-EPRE-Offset cqi-ReportPeriodic OPTIONAL }
spare3, spare2, spare1} -- Need OR INTEGER (-1..6), CQI-ReportPeriodic -- Need ON
CQI-ReportPeriodic ::= CHOICE { release setup cqi-PUCCH-ResourceIndex cqi-pmi-ConfigIndex cqi-FormatIndicatorPeriodic widebandCQI subbandCQI k } }, ri-ConfigIndex OPTIONAL, simultaneousAckNackAndCQI } }
NULL, SEQUENCE { INTEGER (0.. 1185), INTEGER (0..1023), CHOICE { NULL, SEQUENCE { INTEGER (1..4) INTEGER (0..1023) -- Need OR BOOLEAN
-- ASN1STOP The CQI Periodicity and the CQI subframe offset are configured per each UE and signaled to the UE by CQI Configuration Index cqi-pmi-ConfigurationIndex = ICQI/PMI , defined in Table below: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 170/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
cqi-pmiConfigurationIndex = ICQI/PMI
Value of NP
Value of NOFFSET,CQI
ICQI/PMI = 0
1
ICQI/PMI
1 ICQI/PMI 5
5
ICQI/PMI – 1
6 ICQI/PMI 15
10
ICQI/PMI – 6
16 ICQI/PMI 35
20
ICQI/PMI – 16
36 ICQI/PMI 75
40
ICQI/PMI – 36
76 ICQI/PMI 155
80
ICQI/PMI – 76
156 ICQI/PMI 315
160
ICQI/PMI – 156
ICQI/PMI = 316
OFF
n/a
317 ICQI/PMI 511
Reserved
Table 13-11: PUCCH CQI/PMI report configuration Index RI repot periodicity is NP *MRI and RI report subframe is NOFFSET,CQI +NOFFSET,RI, this must be a uplink subframe: ri-ConfigurationIndex = Value of Value of MRI IRI NOFFSET,RI 0 IRI 160
1
−IRI
161 IRI 321
2
− (IRI – 161)
322 IRI 482
4
− (IRI – 322)
483 IRI 643
8
− (IRI – 483)
644 IRI 804
16
− (IRI – 644)
805 IRI 965
32
− (IRI – 805)
IRI = 966
OFF
n/a
967 IRI 1023
Reserved
Table 13-12: PUCCH RI report configuration Index PUCCH resource is given explicitly and maps to RAN1 parameter cqi-PUCCH-ResourceIndex PRB associated with PUCCH resource is given by
(2) nPUCCH ,
(2) m nPUCCH NscRB .
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 171/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
13.2.2.2.2 SR CONFIGURATION The Scheduling Requests are decoded by the eNB since LA3.0. The RRC IEs related to the scheduling request are populated as per the table below: Parameter SchedulingRequestConfigurationFlag
Description Parameter to enable/disable “SchedulingRequestConfiguration” IE
Type
Range
LA3/ LA4 Configuration
Boolean
Enable/Disable
Hardcoded to ‘Enable’
Integer
[0..2047]
pre-calculated in LUTs
Integer
[0..157]
pre-calculated in LUTs
(1)
sr-PUCCHResourceIndex
Parameter nPUCCH,SRI in TS 36.213.
SR configuration index I SR in TS 36.213. Table 13-13: SR configuration parameters sr-ConfigurationIndex
Note: In LA3.0, the value of the sr-PUCCH-ResourceIndex and sr-ConfigurationIndex IEs are precalculated in LUTs. Even though 3GPP Rel’9 have increased the range of sr-ConfigurationIndex, there is no impact on the LUT itself as current release is not supporting the shorter SR periods (1, 2msec) associated with these higher indices. There is no update of TLA3.0 compared with LA3.0, the sr-PUCCH-ResourceIndex and srConfigurationIndex is allocated by LUT, and TLA6.0 inherits the same method/setting of TLA3.0.
13.2.2.2.3 ACK/NACK CONFIGURATION ACK/NAK resource is identified by the combination of length-12 CAZAC sequence and the orthogonal sequences. The ACK/NAK resource used by a UE is determined by Offset for ACK/NAK resource index (1) and Cell-specific cyclic shift value PUCCH signalled by RRC. N PUCCH shift Note: In LA4.0.1, Modem limits the number of regular and SPS A/N supported per PUCCH PRB to 6 and 2 respectively per TTI per cell.
Parameter
Description
Type
Range
n1PUCCH-AN
(1) Parameter N PUCCH in TS 36.213.
Integer
[0..11]
Parameter: PUCCH in shift 36.211, 5.4.1
Enum
deltaPUCCH-Shift
LA3/ LA4 Configuration Pre-calculated (based on cell-id) in LUTs specifically, since LA3.0 (1) N PUCCH nrbSRperTTI .
Pre-configured in LUTs {ds1, ds2, ds3} specifically, since LA3.0, and defaulted as “ds2”.
Table 13-14: PUCCH Format 1a/1b configuration parameters Inherit TLA3.0, The value of n1PUCCH-AN and deltaPUCCH-Shift in TLA6.0 is also pre-calculated in LUTs. According to different configuration scenarios e.g. BW and UL/DL configuration, the value could be different. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 172/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Note that, shortened PUCCH Format 1A/1B is used at every UL subframe since sounding RS is configured at every UL subframe.
13.2.2.2.4 CQI CONFIGURATION There is no update planned for TLA3.0 relative to LA3.0 except that PUCCH format 2a is not supported for TDD because of 3GPP standard constraint. For periodic CQI, only mode 1-0 and 1-1 is supported.
13.2.2.3 PARAMETERS OF PUCCH CHANNEL CONFIGURATION IN TLA3.0/4.X/5.X/6.0 As per LA3.0, the following UE specific parameters is pre-calculated and pre-stored in the LUT for each UE context in TLA3.0. These parameters are related to P-CQI, P-RI, SR, SRS, Measurement Gaps and DRX. Per UE context ID cqi-pmi-ConfigIndex INTEGER (0..1023), cqi-PUCCH-ResourceIndex INTERGER (0.. 1185) ri-ConfigIndex INTEGER (0..1023) sr-PUCCH-ResourceIndex INTEGER (0..2047) sr-ConfigIndex INTEGER (0..155) C-RNTI value INTEGER (11..65523) tpc-RNTI tpc-index
info source in TLA3.0 from LUT (pre-calculated) from LUT (pre-calculated) from LUT (pre-calculated) from LUT (pre-calculated) from LUT (pre-calculated) same as TLA2.x from LUT (pre-calculated) from LUT (pre-calculated) different for all the UEs with the same TPC-RNTI
Table 13-15: Per-UE Parameters from LUTs for TDD Per cell parameters
info source since TLA3.0
n1PUCCH-AN
from LUT (pre-calculated based on cell-id)
nRBCQI pusch-hoppingOffset deltaPUCCH-shift Table 13-16: Per-Cell Parameters from
Parameter
from LUT (pre-configured) from ULS from LUT (pre-configured) MIM and LUTs since TLA3.0
New Per-Cell Parameters to pre-calculate LUTs for TDD Description Type Range TLA3.0 Configuration
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 173/290
LTE Optimization Handbook TLA6.0
cqiInitPeriod
Initial CQI period to be used
MGR/TIPS/NEA
{10sf, 20sf, Enum 40sf, 80sf}
Number of CQIs that can be processed per nbrCQIperTTI TTI due to limited Integer CEM processing power nbrSRperTTI
srsiInitPeriod
nbrSRSperTTI
Parameter srInitPeriod
StartingSROffse t
nbrSRperTTI
Parameter cqiInitPeriod
[5, 100]
Customer Configurable OAM parameter CellL1ULConf::cqiInitPeriod Default value: 20sf Note: Preconfigured in LUTs since TLA3.0 Internal configurable OAM parameter CellL1ULConf::nbrCQIperTTI Default value: 16 Note: Preconfigured in LUTs since TLA3.0
Internal configurable OAM parameter CellL1ULConf::nbrSRperTTI [5, 100] Default value: 16 Note: Preconfigured in LUTs since TLA3.0 Customer configurable OAM parameter {5sf, 10sf, Initial SRS period to CellL1ULConf:: srsInitPeriod Enum 20sf, 40sf, be used Default value: 20sf 80sf} Note: Preconfigured in LUTs since TLA3.0 Internal configurable OAM parameter Maximum number of CellL1ULConf::nbrSRSperTTI SRS that can be Integer [4,16] Default value: 8 multiplexed per TTI Note: Preconfigured in LUTs since TLA3.0 Additional OAM Parameters for SR Configuration for TDD Description Type Range TLA3.0 Configuration MIM Parameter {5sf, 10sf, Initial SR period to Enum 20sf, 40sf, Default value: 20sf be used 80sf} Internal Parameter Starting position of Integer [0,9] Default value: 1 the SR subframe offset allocation Internal Prameter Number of SRs that can be processed per Integer [5, 100] Default value: 16 TTI due to limited CEM processing power Additional OAM Parameters for CQI Configuration for TDD Description Type Range TLA3.0 Configuration MIM Parameter {10sf, 20sf, Enum Default value: 20sf Initial CQI period to 40sf, 80sf} be used MIM Parameter Maximum number of SR that can be Integer multiplexed per TTI
Number of CQIs that nbrCQIperTTI can be processed per Integer TTI due to limited CEM processing power
[5, 100]
Default value: 16
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 174/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Table 13-17: Additional Per-Cell and CQI related parameters to pre-calculate LUTs for TDD
13.2.2.4
ANALYSIS OF PUCCH CONFIGURATION IMPACT ON CAPACITY
The following table shows the maximum capacity per RB according to PUCCH formats. Code Division Multiplexing and Frequency Division Multiplexing is used to multiplex UE’s on the same RB (more accurately RB-pairs) configured for PUCCH resources. PUCCH Format Format 1 Format 1a
UCI information
Maximum Multiplexing Capacity (UE/RB) 36 (deltaPUCCH-Shift=1) 36 (deltaPUCCH-Shift=1)
Scheduling Request (SR) 1-bit HARQ ACK/NACK with/without SR 2-bit HARQ ACK/NACK with/without SR Format 1b (This is for MIMO, 1 bit for each transport 36 (deltaPUCCH-Shift=1) block) Format 2 CQI (20 coded bits) 12 CQI and 1 or 2 bit HARQ ACK/NACK - 20 Format 2 12 bits - Extended CP only CQI and 1 bit HARQ ACK/NACK - (20 + 1 Format 2a 12 coded bits) CQI and 2 bit HARQ ACK/NACK - (20 + Format 2b 12 2 coded bits) Table 13-18: Maximum number of users supported according to PUCCH formats
nRBCQI is a parameter that defines the number of resource blocks for CQI periodic reporting (PUCCH Format 2, 2a or 2b). The maximum capacity of PUCCH CQI could be calculated as: Capacity_PUCCH_CQI = nRBCQI * No. of multiplexed UE per sub-frame * No. of UL subframe in periodicity of P-CQI Assume that, Uplink-downlink configuration1 + per PUCCH capacity for P-CQI =12 (maximum supported users according to 3GPP spec, bCEM processing capability not considered) + 16 ICQI/PMI 35 (i.e. 20ms periodicity) + 322 IRI 482 (i.e. 4*20=80ms periodicity) + nRBCQI=2 (PRB pairs=1), then we can get the maximum number of users that can report CQI periodically in the PUCCH configuration: Capacity_PUCCH_CQI = 1(RB for P-CQI)×12 (UEs per PUCCH)×(2+2)(UL subframes per 10ms period)× 20/10 (number of PUCCH period) (UL subframes)=1×12×8= 96 UEs per CQI reporting period. The calculation of the assumed SRS configuration could be present as below:
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 175/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 13.2-4: PUCCH with P-CQI capacity calculation example The number of UE per PUCCH can be increased by modifying the following parameters: a) Increasing nRBCQI (but the PUSCH capacity will decrease) b) Increasing the periodicity of CQI reporting (increasing CQI-PMIConfigIndex) but this might affect our uplink throughput. c) Increasing the number of UE multiplexed per RB. However, it is up to the operator’s traffic model to decide how to play with these values and achieve the goals planned.
13.3 TLA6.0 PUCCH&SRS CONFIGURATIONS AND RECOMMENDATIONS Considering the system SW capacity and KPI requirement, there are 3 different configuration favorites i.e. operation modes in TLA3.0/4.x/5.x/6.0: maximum user number, maximum performance and maximum KPI. Parameter uplinkControlChannelLUTindex can be used to select different favorite, for each favorite, a set of dedicated UE contexts LUTs are needed. WIPS check is needed to guarantee active user number (maxNbrOfUsers) is in line with uplinkControlChannelLUTindex. Details of each operation mode as below, 15MHz is only supported since TLA6.0. Maximum user number (100AU, 106 UE contexts): uplinkControlChannelLUTindex=0 FRS requires 100 active user numbers per cell. Considering margin for UE re-establishment, the maximum user number for L2 i.e. L2 context number should be 106 in different system bandwidth, UL/DL config and special subframe configurations: - 10MHz+2/8 antenna+UL/DL config 1/2+special subframe 5/7+PRACH format0/4 - 15MHz+2/8 antenna+UL/DL config 1/2+special subframe 5/7+PRACH format0/4 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 176/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
- 20MHz+2/8 antenna+UL/DL config 1/2+special subframe 5/7+PRACH format0/4 In order to support 106 L2 contexts, some configuration parameters of LUT are also changed as table below:
6 6
Init CQI Prd 20 20
Max CQI Prd 20 20
Init SR Prd 20 20
Max SR Prd 20 20
Init SRS Prd 20 20
Max SRS Prd 20 20
#CQI Per TTI 16 16
#SR Per TTI 16 16
#SRS Per SRS Symbol* 8 8
2
8
20
20
20
20
20
16
16
8
UL/DL2+10M
2
8
40
40
80
20
40
16
12**
8
UL/DL2+15M
2
10
40
40
20 40* ** 40
80
20
40
16
12
8
40
16
12
8
nrb CQI
PUCCH PRB
UL/DL1+10M UL/DL1+15M
2 2
UL/DL1+20M
UL/DL2+20M 2 12 40 40 40 80 20 Table 13-19: Uplink control channel/signal LookUpTable-Index0
Note: * SRS number Per SRS symbol: Total SRS UE number per possible SRS symbol including last two symbol of special subframe and last symbol of normal UL subframe. And the value is total SRS UE number including both combs. If SRS number Per SRS symbol =8 then SRS number per comb=4. **In case of UL/DL config 2, up to 4 DL subframe will feedback ACK/NACK in one UL subframe, so the maximum ACK channel number of UL subframe will be 4*NDL, in which NDL is the maximum scheduled DL user per TTI. Current working view is that L1 PUCCH+SR processing capability is 48 per TTI, in order to reserve some channels for SR, per TTI scheduled DL user have to be limited. Assume NDL is 9, then the SR channel number for one TTI is 48-4*9=12. ***For SR period of UL/DL config 2, in order to make all the UE in cell have more faire performance (which is required by CMCC large scale trial), the initial SR period is set as 40ms. There are two possible ways to decrease SR period: -Limit maximum scheduled DL user per TTI NDL, so that more SR channels can be supported in UL TTI. If NDL is 6, then the SR channel number for one TTI is 48-4*6=24 and the initial SR period can be 20ms. -Keep NDL still as 9, but assign more SR resource per TTI (e.g. 24), if ACK+SR total channels is more than L1 processing capability, then L1 should ignore the additional SR from UE without crash.
Maximum performance: uplinkControlChannelLUTindex=1 Besides the maximum active user number scenario, another case is also necessary that all the UE in the cell can get decent performance. Because most case of CMCC large scale trial requires less than 30UE per cell, in order to achieve better performance, the L2 context number will be 32 for all the configurations. Detailed information as below:
UL/DL1+10M UL/DL1+15M
nrb CQI
PUCCH PRB
1 1
4 6
Init/M ax CQI Period 10 10
Init/M ax SR Period 10 10
Init/M ax SRS Period 10 10
#CQI Per TTI 8* 8
#SR Per TTI 8** 8
#SRS Per SRS symbol 8 8
L2 context number 32 32
Active user number 30 30
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 177/290
LTE Optimization Handbook TLA6.0 UL/DL1+20M UL/DL2+10M
1 2
6 8
10 10
UL/DL2+15M
2
10
10
MGR/TIPS/NEA 10 10/20 10/20
10 10
8 16
8 12
8 8
32 32
30 30
10
16
12
8
32
30
12
8
32
30
UL/DL2+20M 2 12 10 10/20 10 16 Table 13-20: Uplink control channel/signal LookUpTable-Index1
Note: *In order to reduce PUCCH PRB number, the number of CQI PRB is set as 1, and then the CQI number per subframe is 8. **In order to reduce PUCCH resource occupation, the number of SR per subframe is also set as 8. Maximum KPI: uplinkControlChannelLUTindex=2 In some test cases, dedicated KPI items are needed e.g. peak UL/DL throughput, beamforming performance and U-plane latency. In order to meet these KPI test requirements, a set of LUTs are designed. During KPI test, there is no user number requirements, so the UE number supported per cell is limited to 3. Detailed information as below. Init/M Init/M Init/Ma #CQI #SR #SRS L2 Active nrb PUCCH ax CQI ax SR x SRS Per Per Per SRS context user CQI PRB* Period Period Period TTI TTI symbol number number UL/DL1+10M
1
2
5
5
5
3
3
3
3
3
UL/DL1+15M
1
4
5
5
5
3
3
3
3
3
UL/DL1+20M
1
4
5
5
5
3
3
3
3
3
UL/DL2+10M
1
4
10
10
5
3
3
3
3
3
UL/DL2+15M
1
4
10
10
5
3
3
3
3
3
UL/DL2+20M
1
4
10
10
5
3
3
3
3
3
Table 13-21: Uplink control channel/signal LookUpTable-Index2 Note: *In order to achieve DL/UL peak throughput, the CFI is fixed as 1 in all cases. So PUCCH PRB is also reduced because of less ACK/NACK channel number. In order to achieve peak throughput, some other parameters should also be changed accordingly e.g. the starting PRB index of RACH message 3, please refer section 5.4.2.1. If the system BW is small e.g. 5MHz and 10MHz, the available CCE number under CFI=1 will be also quite limited. For peak throughput, the C-RNTI should be properly selected to make sure UE can get scheduling chance in each TTI. And according to 3GPP (TS36.213 7.1.7), maximum TBS/cell throughput only occurs in single UE case, so for peak UL/DL throughput test, only one UE should be connected in a single cell. As C-RNTI also related to BW, so different C-RNTI pool is also necessary for Maximum KPI config.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 178/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
14 INTRA LTE MOBILITY OPTIMIZATION HINTS In this chapter, parameters related to mobility are going to be elaborated.. Normally some questions arise in this domain, such as: When to perform Mobility optimization? What method to apply? Which parameters can help improving Mobility?
Usually, the Mobility optimization can occur when the average success rate or average interruption time value doesn’t match the ALU product specification for a determined Bandwidth target. As a basic follow-up, blew is part of the best practice rules, namely
Check the if Neighbours existing Check the all supposed cells /sites are on-air Check that all supposed neighbour relations are declared Evaluate RSSI vs. SNR relation, to see if the neighbours are matching expectations based on coverage map. Perform the drive test to see whether our mobility-related parameters are of its best value.
If we could guarantee that these values are “normal”, the chances to have performance issues are much less difficult to occur. If there are still some performance issues regardless of the correct configurations, then below parameters can be used in order to correct the situation.
14.1 NEIGHBOUR CONFIGURATION Just as described above, the first thing that should be confirmed is to declare neighbouring relationship correctly. One of the ways to declare the best neighbour relation is working with the information done by ARFCC team previously. With Cell ID coverage we can check the cell ID’s neighbour’s for each sector as showing bellow. Drive test and post processing the measurements should then be done to confirm the neighbour list and the HO success rate between sites.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 179/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.1-1: Coverage (Best Cell ID) analysis of South Pudong road area in CMCC LST – done by ARFCC team Other possible way to perform the neighbour list is to run ANR feature in the network. The feature will create the X2 links that would be need for each sector. Maximum of 32 X2 visible links could be created. Be aware the feature should be activated after the first neighbour optimization performed and after the users start to perform data traffic. Here the X2-Link is created but the name of the neighbour comes in Binary. Below depicts that site NE0054 create X2 link with site NE0034 and in the Wips it appears as X2Access/0_330_110_NE0054_11.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 180/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.1-2: X2-Link access
14.1.1 INTRA FREQUENCY NEIGHOUR DECLARATION In the beginning when network is deployed, inside CIQ is necessary to fulfil the information relative to the Intra Neighbouring cell’s of each sector. After network deployment a first check in the neighbour list should be done, to ensure the intra HO neighbour’s are correctly created. Example bellow in a) Intra eNB Neighbour cell example b) Inter eNB Neighbour cell example Figure 14.1-3 shows the sector LST001_1 have Intra neighbour with LST001_2 and LST001_3, etc.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 181/290
LTE Optimization Handbook TLA6.0
a) Intra eNB Neighbour cell example
MGR/TIPS/NEA
b) Inter eNB Neighbour cell example
Figure 14.1-3: LTE Intra Frequency Neighbour list
14.1.2 INTER FREQUENCY NEIGHBOUR DECLARATION Example below in Figure 14.1-4 shows the sector LST008_1 have Intra neighbours of sector LST008_2 and LST008_3, etc.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 182/290
LTE Optimization Handbook TLA6.0
a) Intra eNB Neighbour cell example
MGR/TIPS/NEA
b) Inter eNB Neighbour cell example
Figure 14.1-5: LTE Inter Frequency Neighbour list
14.2 MOBILITY PARAMETERS Some parameters are the base to optimize the mobility. In some of then we are able to change and get better results, for example in terms of performance (HO Success Rate). Note for different networks and environments the recommended values could not be the best to apply. Always run a first drive and after with recommended parameters try to setup the best for your network. Object
Name
Recommended Value
CellSelectionReselectionConf
qRxLevMin
-120
CellSelectionReselectionConf
qRxlevminoffset
2
LteNeighboringCellRelation
threshXLow
0
CellSelectionReselectionConf
qHyst
dB2
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 183/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
RrcMeasurementConf
filterCoefficientRSRP
Fc8
ReportConfigEUTRA
Hysteresis
2
ReportConfigEUTRA
timeToTrigger
ms40
LteNeighboringCellRelation
cellIndividualOffset
dB0
Table 14-1: Some Mobility parameters
14.2.1 MOBILITY IN RRC IDLE MODE Cell Reselection is a procedure triggered by the UE in Idle Mode to determine which LTE cell to camp on. The trigger can be internal (e.g. periodic trigger to ensure that UE is still on the best cell) or external (e.g. upon change of Cell Reselection parameters broadcast on the selected cell’s BCH). The cell selection and reselection is controlled by the System Information parameters provided in SIB1, SIB3, SIB4 and SIB5.
14.2.1.1
CELL SELECTION & RESELECTION
Cell Reselection is a procedure triggered by UEs in Idle Mode to determine which LTE cell to camp on. When a UE, camps on a cell it monitors its broadcast and paging channels. The procedure is internal to the UE and there is therefore no EUTRA level use case for it. According to 3GPP, intra-LTE selection occurs when the S criterion is satisfied, namely
Srxlev > 0 where
Srxlev = Qrxlevmeas – (Qrxlevmin + Qrxlevminoffset) - Pcompensation Once the selection is accomplished, UE dominated reselection is about to happen. In order to further restrict the amount of measurement carried out by the UE in RRC-Idle mode, the following rules are used by the UE: IF SServingCell > Sintrasearch -> UE may choose to not perform intra-frequency measurements If SServingCell ≤ Sintrasearch -> UE shall perform intra-frequency measurements And If SServingCell > Snonintrasearch, UE may choose to not perform inter-frequency measurements of the equal priority and lower priority network. If SServingCell ≤ Snonintrasearch, UE shall perform inter-frequency measurements of the equal priority and lower priority network. For intra-frequency and equal priority inter frequency cell reselection, R criterion is used: i.e., Rs < Rn where
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 184/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Rs = Qmeas,s + QHyst Rn = Qmeas,n - Qoffset UE will reselect the new first cell in the ranked list based on criterion above if both below conditions are met: The new cell is better ranked than the serving-cell during a time interval tReselectionEUTRAN More than 1 second(s) has elapsed since the UE camped on the current serving-cell Measurement phase RSRP > Qrxlevmin + Qrxlevminoffset + Sintrasearch ≤ Qrxlevmin + Qrxlevminoffset + Sintrasearch Measurement phase
Figure 14.2-1: LTE to LTE Mobility – Measurement phase (RSRP vs. Time)
Ranking phase RSRP
RSRP(n)qOffset
Ranked list
neighbour-decision
qHyst
3
Rs
-100
3
qOffSet
3
Rn1
-95
-98
4
Rn2
-94
-97
3
Rn3
-93
-96
2
Rn3
Rn4
-92
-95
1
Rn4
Figure 14.2-2: LTE to LTE Mobility – Ranking Phase For example in
Figure 14.2-2, UE shall rank the measured cells based on the above ranking criterion, as below: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 185/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Rs for serving cell: Rs = Qmean,s+qHyst = -100+3 =-97dBm ==> Rank3 Rn for neighbour cell1: Rn1 = Qmean,n-qOffset = -95-3 = -98dBm ==> Rank4 Rn for neighbour cell2: Rn2 = Qmean,n-qOffset = -94-3 = -97dBm ==> Rank3 Rn for neighbour cell3: Rn3 = Qmean,n-qOffset = -93-3 = -96dBm ==> Rank2 Rn for neighbour cell4: Rn4 = Qmean,n-qOffset = -92-3 = -95dBm ==> Rank1 Then UE select the suitable cell according to the ranked list, i.e. UE select the first ranked neighbour cell4 (Rn4). Decision phase
RSRPn4
RSRPs
QoffSet qHyst
Rn4 reselection
tReselectionEUTRAN Figure 14.2-3: LTE to LTE Mobility – Decision Phase (RSRP vs. Time) For inter frequency and also inter-RAT frequency with different priority reselection, we have: Criteria 1: the SnonServingCell,x of a cell on evaluated frequency is greater than ThreshXHigh during a time interval TreselectionRAT; Cell reselection to a cell on a higher priority E-UTRAN frequency or inter-RAT frequency than serving frequency shall be performed if: -
A cell of a higher priority E-UTRAN frequency or inter-RAT frequency fulfils criteria 1; and
-
More than 1 second has elapsed since the UE camped on the current serving cell.
Cell reselection to a cell on a lower priority E-UTRAN frequency or inter-RAT frequency than serving frequency shall be performed if: No cell on a higher priority E-UTRAN frequency or inter-RAT frequency than serving frequency fulfills the criteria 1; and No cell on serving frequency or on an equal priority E-UTRAN frequency fulfills the criteria in section 14.2.1.1 (i.e. R criterion); and SServingCell < Threshserving, low and the SnonServingCell,x of a cell of a lower priority E-UTRAN frequency or inter-RAT frequency is greater than ThreshXLow during a time interval TreselectionRAT; and -
more than 1 second has elapsed since the UE camped on the current serving cell.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 186/290
LTE Optimization Handbook TLA6.0
RSRP
MGR/TIPS/NEA
UE starts inter-freq meas for cell reselection LTE Serving Cell 4
Target cell is reselected
1 Qrxlevmin(SIB3)+ sNonIntraSearch
tReselectionEUTRAN
RSRP LTE Target Cell Qrxlevmin+Qrxlevminoffset +Pcompensation +threshXHigh
3 Qrxlevmin(SIB3)+ threshServingLow
Qrxlevmin+Qrxlevminoffset +Pcompensation +threshXLow
2
3 time Figure 14.2-4: LTE to LTE Mobility – Measurement phase (RSRP vs. Time) For further details, please refer to TDD Lte paramters user guide.
14.2.1.2
QRXLEVMIN
Clarifications regarding qRxLevMin: A parameter with this name appear in several objects and is then transmitted to UE inside several system information block types i.e. Sibs: CellSelectionReselectionConf – transmitted in SIB1 and SIB3 CellReselectionConfUtraFdd – transmitted in SIB6 (see chapter 15) CellReselectionConfUtraTdd – transmitted in SIB6 (see chapter 15) CellReselectionConfGERAN – transmitted in SIB7 (see chapter 15) The LTE – GERAN mobility is using two of them, the one sent in SIB3 and the one sent in SIB7. The IE SystemInformationBlockType3 contains cell re-selection information common for intrafrequency, inter-frequency and/or inter-RAT cell re-selection (i.e. applicable for more than one type of cell re-selection but not necessarily all) as well as intra-frequency cell re-selection information other than neighbouring cell related. This parameter configures the cell min required RSRP level used by the UE in cell reselection. Recommended & Default Value= "-120" Expected behaviour when changing this parameter Increasing the value of this parameter would: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 187/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Determine the UE to start cell reselection sooner and then will artificially decrease cell size in idle mode. Decreasing the value of this parameter would: Determine the UE to start cell reselection later and then will artificially increase cell size in idle mode. KPI Impact: Mobility - low values might create coverage discontinuity in idle, as seen by UE. The optimization process should contain the following steps: Step 1: Set the value of qRxLevMin to one of the following values {- 124, -122, -120, -118, -116}. Step 2: With UE in idle mode, perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qRxLevMin and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.3
SINTRASEARCH
This parameter specifies the threshold for the serving cell reception level, below which the UE triggers intra-frequency measurements for cell reselection.
Recommended & Default Value= "62" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an early start of measurement for cell reselection. Decreasing the value of this parameter would: Determine a late start of measurement for cell reselection. KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection. The optimization process: test results indicate sIntraSearch should be set to the highest allowed value to minimize SINR degradation in reselection boundaries. Although as an experiment process; values such as 30 or 40 for the sIntraSearch might be used.
14.2.1.4
SNONINTRASEARCH
This parameter is used for setting a threshold for the selection criterion, threshold that would determine when, based in serving cell field level, the UE starts performing measurements for interfrequency and inter-RAT measurements. It is used for cell reselection. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 188/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "16" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start earlier the measurement for inter-frequency reselection which will probably empty the UE battery sooner. Decreasing the value of this parameter would: Determine the UE to start later the measurements for inter-frequency reselection. Possible impact correct and timely reselection for high speed UEs. KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection. The optimization process should contain the following steps: Step 1: Set the value of sNonIntraSearch to one of the following values {12, 13, 14, 15, and 16}. Step 2: With UE in idle mode, perform a drive test back and forth between the cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another sNonIntraSearch and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.5
QHYST
This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Broadcast in SystemInformationBlockType3. Recommended & Default Value= "dB2" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a later reselection of the target neighbouring cell (smaller the target cell list). Decreasing the value of this parameter would: Determine an earlier reselection of the target neighbouring cell (larger the target cell list). KPI Impact: Mobility - can improve mobility by determining timely reselection. The optimization process should contain the following steps: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 189/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 1: Set the value of qHyst to one of the following values {dB1, dB2, dB3, dB4, dB5}. Step 2: With UE in idle mode, perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHyst and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.6
QOFFSETCELL
This parameter defines the offset between the current LteCell and the LteNeighboringCell. Recommended & Default Value= "dB0" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a later reselection of the target neighbouring cell (larger target cell list). Decreasing the value of this parameter would: Determine an earlier reselection of the target neighbouring cell (smaller target cell list).
KPI Impact: Mobility - higher the value later the cell reselection. Lower the value, earlier the cell reselection.
The optimization process should contain the following steps: Step 1: Set the value of qOffsetCell to one of the following values {dB1,dB2,dB3,dB4,dB5,dB6,dB7}. Step 2: With UE in idle mode, perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qOffsetCell and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.7
QRXLEVMINOFFSET
This parameter defines an offset to be applied in cell selection criteria by the UE when it is engaged in a periodic search for a higher priority PLMN. Recommended & Default Value= "2" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier reselection of the target neighbouring cell. Decreasing the value of this parameter would: Determine a later reselection of the target neighbouring cell. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 190/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values might create coverage discontinuity during selection operation due to shrinking target cell as seen by the UE. Note: If you do not use inter-PLMN mobility, this parameter is inhibited.
14.2.1.8
THRESHSERVINGLOW
This threshold for serving cell reception level is used in reselection evaluation towards lower priority E-UTRAN frequency or RAT. The value sent over the RRC interface is half the value configured (the UE then multiplies the received value by 2) Defined in TS 36.331 Broadcast in SystemInformationBlockType3 The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionEUTRAN. The parameter discussed here only impacts the part related to the serving cell. The condition on the serving cell can be rewritten as a condition on the measured level in the serving cell as follows:
Q relevmeas qRxLevMin threshServ ingLow The optimization of threshServingLow is based on this relation. Recommended & Default Value= "0" NEA Recommended Value This value depends on the strategy of the operator and the network coverage. In the inter-frequency mobility scenario, the value can be set higher to make UE easier to reselect to lower priority frequency Neighbour cell. Note: To get the option to reselect as soon as possible, with serving cell reception level below sNonIntraSearch, we can set threshServingLow at the same level than sNonIntraSearch. i.e. set threshServingLow to 16 dB. Expected behaviour is as follows when changing this parameter Increasing the value of this parameter could: Determine an earlier selection of target cell, i.e. a shrinking of the serving cell in idle mode. Decreasing the value of this parameter would: Determine a later selection of target cell which is similar to a shrinking of the target cell.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 191/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Optimization of this parameter, in conjunction with threshXLow should aim at obtaining the cell sizes for both target cell and serving cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshServingLow should be decoupled from the optimization for threshXLow. For this, the value of threshXLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshServingLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshServingLow to one of the following values {8, 9, 11, 12, and 13}. Step 2: With UE in idle mode, perform a drive test back and forth between the cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshServingLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.9
THRESHXLOW
This parameter sets the threshold of the selection criteria towards lower priority E-UTRAN frequency or RAT. The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionEUTRAN. The parameter discussed here only impacts the part related to the serving cell. The condition on the serving cell can be rewritten as a condition on the measured level in the serving cell as follows:
Qrelevmeas > Qrxlevmin + Pcompensation + threshXLow The optimization of threshXLow is based on this relation.
Recommended & Default Value= "0"
NEA Recommended Value: This value depends on the strategy of the operator and the network coverage. Expected behaviour is as follows when changing this parameter Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 192/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Increasing the value of this parameter could: Determine a later selection of target cell which is similar to a shrinking of the target cell. Decreasing the value of this parameter would: Determine an earlier selection of target cell, i.e. a shrinking of the serving cell in idle mode.
Optimization of this parameter, in conjunction with threshServingLow should aim at obtaining the cell sizes for both target cell and serving cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshXLow should be decoupled from the optimization for threshServingLow. For this, the value of threshServingLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshXLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshXLow to one of the following values {0, 6, 12, 18, and 24}. Step 2: With UE in idle mode, perform a drive test back and forth between the cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshXLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.10 THRESHXHIGH This parameter sets the threshold of the selection criteria towards lower priority E-UTRAN frequency or RAT. The condition on the serving cell can be rewritten as a condition on the measured level in the serving cell as follows:
Qrelevmeas > Qrxlevmin + Pcompensation + threshXHigh The optimization of threshXHigh is based on this relation. Recommended & Default Value= "10"
Expected behaviour when changing this parameter Increasing the value of this parameter could: Determine a later selection of target cell which is similar to a shrinking of the target cell. Decreasing the value of this parameter would: Determine an earlier selection of target cell, i.e. a shrinking of the serving cell in idle mode. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 193/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 1: Set the value of threshXHigh to one of the following values {0, 6, 10, 14, 16, 18 and 22}. Step 2: With UE in idle mode, perform a drive test back and forth between the cells on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshXHigh and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the target cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate
14.2.1.11 TRESELECTIONEUTRAN This parameter specifies the value of the cell reselection UE timer in the serving cell. Broadcast in SystemInformationBlockType3. It imposes a condition on the reselection. UE will actually reselect the new cell, only if the new cell is better ranked than the serving cell during a time interval tReselectionEUTRAN.
Recommended & Default Value= "2" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a delayed reselection which could be an issue for fast moving UEs. Decreasing the value of this parameter would: Facilitate ping-pong behaviour during reselection process. KPI Impact: Mobility - low values of this parameter might allow ping-pong behaviour during reselection operation. High values of this parameter might delay the reselection and possible lead to lost connection to the serving cell. Optimization of this parameter should find a trade-off between delayed reselection and ping pong behaviour. Most probably, if the UEs are not moving fast, the delayed reselection would not be an issue. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionEUTRAN to one of the following values {1, 2, 3, 4}. Step 2: With the UE in idle mode, perform a drive back and forth between the cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionEUTRAN. Step 4: Post-process the logs and analyze them as reselection position vs. tReselectionEUTRAN values and ping pong behaviour vs. tReselectionEUTRAN values and choose the optimized value to obtain smallest interruption time and highest success rate. Step 5: Calculate the HO success rate in each direction.
14.2.1.12 TRESELECTIONEUTRASFMEDIUM This parameter configures the t-ReselectionEUTRA-SF included in the IE SystemInformationBlockType3. Parameter “Speed dependent ScalingFactor for tReselectionEUTRAN” (TS 36.304). If the field is not present, the UE behaviour is specified in TS Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 194/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
36.304. The concerned mobility control related parameter is multiplied with this factor if the UE is in Medium Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended Value = “oDot5”. Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionEutraSfMedium to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionEutraSfMedium. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.13 TRESELECTIONEUTRASFHIGH This parameter configures the t-ReselectionEUTRA-SF included in the IE SystemInformationBlockType3. Parameter “Speed dependent ScalingFactor for tReselectionEUTRAN” (TS 36.304). If the field is not present, the UE behaviour is specified in TS 36.304. The concerned mobility control related parameter is multiplied with this factor if the UE is in High Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended & Default Value= "oDot25" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 195/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionEutraSfHigh to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the cells on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionEutraSfHigh. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.14 TEVALUATION This parameter configures the duration for evaluating criteria to enter mobility states
Recommended & Daefault Value= "s30" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE spend more time evaluating the criteria to enter mobility state; meaning a handover delay. Decreasing the value of this parameter would: Determine the UE spend more time evaluating the criteria to enter mobility state; meaning a handover speed up (to be noted that the smaller value allowed is already in used s30). KPI Impact: Mobility - low values of this parameter will determine the UE to spend less time evaluating criteria for enter mobilityr. This results in a speed up of the handover process. Higher values of this parameter will determine the UE to spend more time evaluating criteria for enter mobilityr. This results in a delay of the handover process A procedure that optimizes tEvaluation would contain the following steps: Step 1: Set the value of tEvaluation to one of the following values {s30, s60, s120, s180, and s240}. Step 2: Perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another tEvaluation and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.15 NCELLCHANGEHIGH This parameter configures the number of cell changes to enter high mobility state
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 196/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "12" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. A procedure that optimizes nCellChangeHigh would contain the following steps: Step 1: Set the value of nCellChangeHigh to one of the following values {10, 11, 12, 13, and 14}. Step 2: Perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.16 NCELLCHANGEMEDIUM This parameter configures the number of cell changes to enter medium mobility state Recommended & Default Value= "4" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility – low values of this parameter might allow the UE to start cell reselection earlier. High values of this parameter might allow the UE to start cell reselection later. A procedure that optimizes nCellChangeMedium would contain the following steps: Step 1: Set the value of nCellChangeMedium to one of the following values {1, 2, 3, 4, 5, 6}. Step 2: Perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 197/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
14.2.1.17 QHYSTSFHIGH This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-High included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in High Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility – low values of this parameter might allow the UE to start cell reselection earlier. High values of this parameter might allow the UE to start cell reselection later. A procedure that optimizes qHystSfHigh would contain the following steps: Step 1: Set the value of qHystSfHigh to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.1.18 QHYSTSFMEDIUM This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-Medium included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in Medium Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 198/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility – low values of this parameter might allow the UE to start cell reselection earlier. High values of this parameter might allow the UE to start cell reselection later. A procedure that optimizes qHystSfMedium would contain the following steps: Step 1: Set the value of qHystSfMedium to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2 MOBILITY IN RRC CONNECTED MODE
Figure 14.2-5: LTE to LTE Mobility – Handover cases Event A3 – Neighbour becomes offset better than serving Entering condition for this event:
Mn = measurement result of the neighbouring cell [dBm] Ofn = MeasObjectEUTRA::offsetFreq, corresponding to the neighbouring cell [dB] Ocn = cellIndividualOffset for neighbouring cell [dB] Hys = reportConfigEUTRA::hysteresis [dB] Ms = measurement result of the serving cell [dBm] Ofs = MeasObjectEUTRA::offsetFreq, corresponding to the serving cell [dB] Ocs = cellIndividualOffset for serving cell [dB] Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 199/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Off = eventA3Offset [dB] Leaving condition for this event:
For TDD, event A3 is used as well for inter-frequency handover. So the Event A5 is only used for FDD. Event A5: when the serving cell becomes worse than a given threshold and the neighbour cell becomes better than a given absolute threshold2 Entering condition for this event:
&
14.2.2.1
FILTERCOEFFICIENTRSRP
This parameter configures the RRC IE filterCoefficientRSRP included in the IE quantityConfigEUTRA in the MeasurementConfiguration IE. If this parameter is not configured (absent) then the default RRC value defined in 36.331 is used by the eNB and signalled to the UE. The RSRP values reported by the UE are obtained by filtering several measurements performed by the UE. If this filter can allow quick variation to be reported or it can rely more on the last reported value and less on the measured value such that there is less variation in the sequence of the reported value. The higher the value of filterCoefficientRSRP the smoother the reported measurement will be and consequently the less likely ping-ponging occurs between sectors during handover.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 200/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-6: Theoretical view -70 1
106 211 316 421 526 631 736 841 946 1051 1156 1261 1366 1471 1576 1681 1786 1891
-80
-90 RSRP_instant -100
RSRP_FC(K=4) RSRP_FC(K=11)
-110
-120
-130
Figure 14.2-7: filterCoefficientRSRP - Theoretical comparison (Simulation Analysis)
Recommended & Default Value= "fc8" Expected behaviour when changing this parameter: Increasing the value of this parameter would: Decrease the variation in the reported RSRP value. Decrease ping-pong between the cells in case of handover conditions. Delay the speed at which the reported RSRP adapts to the RSRP variation. This might eventually slightly delay the HO, if the value of the parameter is too high. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 201/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Improve the system behaviour regarding the throughput during HO. Decreasing the value of this parameter would: Increase the variation in the reported RSRP value due to noise. Increase the ping-pong between the cells in case of handover conditions due to variations in reported RSRP. Decrease the HO quality relative to throughput. Increase the speed at which the reported RSRP adapts to the RSRP variation. This might speed up the HO which could manifest as ping-pong.
KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Optimization of this parameter should be performed in conjunction with optimization of hysteresis and timeToTrigger parameters. Finding the optimum pair of (filterCoefficientRSRP, hysteresis, and timeToTrigger) should consider the following steps: Step 1: Set the values of filterCoefficientRSRP and to hysteresis and to timeToTrigger to one of the following {(fc6,2,80), (fc8,3,40), (fc8,4,20), (fc5,1,100)}, in both current cell and neighbour cell. Step 2: Perform a drive test while performing a download and log the throughput values and the position of the UE. Drive in and out of the current cell to the neighbour cell. Step 3: Repeat Step 2 for another pair of values of the three tested parameters. Step 4: Represent throughput vs. position (distance) (Service continuity), #HO-attempts, Success Rate/Failure Rate, #of Ping-pongs, HO interruption time for all pairs of tested values. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.2
HYSTERESIS
This parameter configures the RRC IE hysteresis included in the IE reportConfigEUTRA in the MeasurementConfiguration IE. This parameter defines the hysteresis used by the UE to trigger an intra-frequency event-triggered measurement report. It is used in several processes: Event A1 (Serving becomes better than threshold); Event A2 (Serving becomes worse than threshold); Event A3 (Neighbour becomes offset better than serving); Event A4 (Neighbour becomes better than threshold); Event A5 (Serving becomes worse than threshold1 and neighbour becomes better than threshold2. Recommended & Default Value= "2" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO due to the more important difference that must exist between the serving cell and neighbour cell. Drop the call if the value is too large i.e. connection to the serving cell is lost before having reached the neighbour cell level that satisfies the HO condition. Decreasing the value of this parameter would: Create a ping – pong behaviour because the measurement quick variations (noise-like) might trigger HO decisions. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 202/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Throughput - low values of this parameter can generate a ping pong behaviour which can result in interruption times and low throughput during HO operation. Optimization of this parameter should be performed in conjunction with optimization of filterCoefficientRSRP and timeToTrigger parameters, as presented in the previous paragraph.
14.2.2.3
TIMETOTRIGGER
This parameter sets the time duration time during which the conditions to trigger an event report have to be satisfied before sending a RRC measurement report in event triggered mode. Recommended & Default Value= "ms40" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO decision. Determine a call drop due to significant serving cell signal degradation before timeToTrigger expires. Decreasing the value of this parameter would: Generate ping-pong HO behaviour due to the fact that quick variations of the measured signal (noise-like variations) might satisfy the HO relation for the short while represented by timeToTrigger but not much longer.
KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell.
This parameter should be carefully optimized, best in conjunction with filterCoefficientRSRP and hysteresis as presented in paragraph 14.2.2.1 and 14.2.2.2. Indeed, the optimized value can be impacted by the load of the surrounding cells.
14.2.2.4
CELLINDIVIDUALOFFSET
This parameter defines the cell individual offset between the current LteCell and the neighbour cell provided to the UE in RRC Connected mode for measurement.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 203/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "dB-3" Expected behaviour when changing this parameter Increasing the value of this parameter would: Increase the number of ping pongs between the 2 cells and speed up the HO. Decreasing the value of this parameter would: Decrease the number of ping pongs between the 2 cells and delay the HO.
KPI Impact: Mobility – low values of this parameter will delay the HO, and high values will generate pingpong behaviour.
The following steps could be used if you plan to optimize this parameter: Step 1: If you detect a consist dropping from a specific cell, you can think to tune the cellIndividualOffset in steps of dB2 units. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of cellIndividualOffset. Step 4: Post process the measurement and analyze them as HO success rate vs. cellIndividualOffset values and ping pong behaviour vs. cellIndividualOffset values and choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.5
EVENTA3OFFSET
This parameter is used to indicate an event (A3) specific offset of the serving frequency to be applied when evaluating triggering conditions for measurement reporting. Recommended & Default Value = “2” Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to handover to the target cell later Decreasing the value of this parameter would: Determine the UE to handover to the target cell earlier.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 204/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility – low values of this parameter might allow the UE to handover earlier; while high values might delay the handover
The following steps could be used if you plan to optimize this parameter: Step 1: Set the values of eventA3offset to one of the following values {-3,-2,-1, 0, 1, 2, 3}. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of eventA3Offset Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.6
OFFSETFREQ
This parameter is used to indicate a frequency specific offset to be applied when evaluating triggering conditions for measurement reporting. Recommended & Default Value= "dB0" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to find the target cell earlier, which is similar with a shrinking of the serving cell. Decreasing the value of this parameter would: Determine the UE to find the target cell later, which is similar with a shrinking of the target cell. KPI Impact: Mobility – low values of this parameter might allow the UE to determine the strongest cell later. High values of this parameter might allow the UE to determine the strongest cell earlier.
The following steps could be used if you plan to optimize this parameter: Step 1: Set the values of offsetFreq to one of the following values {-3,-2,-1, 0, 1, 2, 3}. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of offsetFreq. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.7
THRESHOLDEUTRARSRP
This parameter configures the first threshold to be used for event A5 measurement reporting. Recommended & Default Value= "-100" Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 205/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of the target neighbouring cell. Decreasing the value of this parameter would: Determine a later selection of the target neighbouring cell.
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of thresholdEutraRsrp to one of the following values {-104, -102, -100, -98, -96}. Step 2: Perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another thresholdEutraRsrp and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the new cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.8
THRESHOLD2EUTRARSRP
This parameter configures the second threshold to be used for event A5 measurement reporting. This parameter configures the RRC IE Threshold EUTRA RSRP included in the IE reportConfigEUTRA in the MeasurementConfiguration IE. This IE should be present if the parameter triggerTypeEUTRA is set to eventA1, eventA2, eventA4 or eventA5 and triggerQuantity is set to RSRP. Otherwise it should be absent. Recommended & Default Value= "-100" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a later selection of the target neighbouring cell. Decreasing the value of this parameter would: Determine an earlier selection of the target neighbouring cell.
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of threshold2EutraRsrp to one of the following values {-104, -102, -100, -98, -96}. Step 2: Perform a drive test back and forth on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 206/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 3: Choose another threshold2EutraRsrp and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the new cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.9
REPORTINTERVAL
This parameter configures the RRC IE reportInterval included in the IE reportConfigEUTRA in the MeasurementConfiguration IE. The ReportInterval indicates the interval between periodical reports. The ReportInterval is applicable if the UE performs periodical reporting (i.e. when reportAmount exceeds 1), for triggerType ‘event’ as well as for triggerType ‘periodical’. Recommended & Default Value= "ms240" Expected behaviour when changing this parameter Increasing or decreasing the value of this parameter would: Will not decrease or increase the Handover Success rate; but no point to have high values.
KPI Impact: Mobility – No specific issues using higher or lower values; a compromise needs to be found. A procedure that optimizes reportInterval would contain the following steps: Step 1: Set the value of reportInterval to one of the following values {120, 240, 480, 640, 1024, and 2048}. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportInterval. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.10 MAXREPORTCELLS This parameter defines the maximum number of cells to be reported in a measurement report. Recommended & Default Value= "1"; if ANR active it should be set to "8" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to report as many new neighbour cells as possible in a short time. Decreasing the value of this parameter would: Determine the UE to report fewer neighbour cells. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells.. A procedure that optimizes maxReportCells would contain the following steps: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 207/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 1: Set the value of maxReportCells to one of the following values {1, 2, 3, 4, 5, 6, 7, and 8}. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of maxReportCells. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
14.2.2.11 REPORTAMOUNT This parameter configures the number of periodical reports the UE has to transmit after the event was triggered. Recommended & Default Value= "r8" Expected behaviour when changing this parameter Increasing the value of this parameter would: Increase the Handover Success Rate for multiple repetitions in bad RF conditions. Decreasing the value of this parameter would: Decrease the Handover Success Rate for multiple repetitions in bad RF conditions. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells.. A procedure that optimizes reportAmount would contain the following steps: Step 1: Set the value of reportAmount to one of the following values {r1, r2, r4, r8, r16, r32, r64}. Step 2: Perform a drive test in and out of the cell and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportAmount. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 208/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
14.2.2.12 CALL FLOW FOR INTER-ENB MOBILITY, X2 HO – UE IN RRC CONNECTED
Figure 14.2-8: Call flow for Inter-eNB mobility, X2 HO – UE in RRC Connected (1)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 209/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-9: Call flow for Inter-eNB mobility, X2 HO – UE in RRC Connected (2)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 210/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
14.2.2.13 CALL FLOW FOR INTER-ENB MOBILITY, S1 HO – UE IN RRC CONNECTED
Figure 14.2-10: Call flow for Inter-eNB mobility, S1 HO – UE in RRC Connected (1)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 211/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-11: Call flow for Inter-eNB mobility, S1 HO – UE in RRC Connected (2)
14.2.2.14 INTRA LTE HANDOVER OPTIMIZATION EXAMPLES IN FIELD TEST Handover success rate is very import KPI for wireless network, low handover success rate will lead low throughput, high data interrupt latency and bad experience to customer. Hereunder are some handover optimization examples got from CMCC LST field test.
14.2.2.14.1
COVERAGE ISSUE
We met this issue at site Yuewei and Shangpi in CMCC LST south Pudong road area according to Figure 14.2-12.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 212/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-12: Test route and RSRP before optimization
Figure 14.2-13: UE Signalling and Events in target cell before optimization From Figure 14.2-15, we found that when UE moved from cell_108 (cell1 of site Shangpi) to cell_121 (cell2 of site Yuewei), there were 2 measurement reports reported by UE and indicated target cell fulfilled A3 event, but no handover occurred and UE hanged on source cell till UE became to RRC Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 213/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
idle, then UE failed to initial RRC reestablishment procedure because T301 expired. From Figure 14.2-12, we also found that the RSRP in most position of the route between these two eNB were below -106dBm, which means the coverage here was poor. To solve the issue, we did steps as below: Step1: Checked the LteNeighboringFreqConf and LteNeighboringCellRelation of the two eNBs, verified that there is no error of LteNeighboring configuration in both eNBs; Step2: Because UE reported the right measurement result, the UeMeasurementConf and intra LTE mobility error could be excluded, but the pusch power reached 23dBm and UE became RRC idle, which means UE low layer may reported radio link failure to its’ high layer. We cannot get detail traces between every layers of UE, so, we captured callp trace in source eNB and found that eNB did not receive the measurement report, which lead to the issue. Step3: Because the coverage was poor and UE may reported radio link failure, we modified the downtitle of cell_108 from 0˚ to 3˚, cell_121from 0˚ to 2˚ to try to improve the coverage and tested again to check the issue was solved. The optimization results please refer to chapter 14.2.2.14.4.
14.2.2.14.2 14.2.2.14.2.1
MEASCONFIG ISSUES
INTRA LTE MOBILITY DISABLED ISSUE
We met this issue at site Lanchun and Pujiancai in CMCC LST south Pudong road area according to Figure 14.2-14.
Figure 14.2-14: Test route and RSRP before optimization
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 214/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-15: UE Signalling and Events in target cell before optimization
UE did not receive measConfig in source cell
Figure 14.2-16: No measconfig received when UE attach in source cell before optimization From Figure 14.2-15, we found that when UE moved from cell_89 (cell3 of site Lanchun) to cell_100 (cell2 of site Pujiancai), there was no measurement report even the signal level of target cell became fulfil the trigger condition of event A3, which lead to no handover occurred and UE hanged on source cell till UE became to RRC idle, then UE initial RRC reestablishment procedure after camp on target cell. Based on Figure 14.2-16, we checked the signalling message of source cell, and found that there is no measConfig received in source cell which means UE will not perform measurement for intra/inter frequency, that’s the reason of the “No MR”. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 215/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
To solve the issue, we did steps as below: Step1: Checked the UeMeasurementConf of the two eNBs, verified that there is no error of UE measurement configuration in both eNBs; Step2: Because the whole network is using the same EARFCN, so we checked the value of isIntraFreqMobilityAllowed (mobility enable/disable) parameter in source eNB, and found that the value is “false”, which lead to the issue. Step3: Modified the value of isIntraFreqMobilityAllowed to “true” and tested again to check the issue was solved.
Figure 14.2-17: RSRP became better after optimization
Figure 14.2-18: UE successfully HO to target cell after optimization
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 216/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
UE received measConfig from eNB
Figure 14.2-19: Detail of measConfig message received after optimization After the optimization, we found the measConfig received by UE as Figure 14.2-19, and cell_89 could successfully handover to cell_100, the RSRP in the test route became better too.
14.2.2.14.2.2
MEASURED CARRIER FREQUENCY ERROR ISSUE
We met this issue at site Keyuan and Chunxiao in CMCC LST Zhangjiang area according to Figure 14.2-20.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 217/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-20: Test route and RSRP before optimization
No neighbour cell measured Figure 14.2-21: UE Signalling and Events in source cell before optimization
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 218/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
UE received measConfig in source cell
Measured carrierFreq was set to error value in measConfig
Figure 14.2-22: Detail signalling message when no measurement report before optimization From Figure 14.2-21, we found that there was no measurement report and no handover when UE went from cell_29 (cell3 of site Keyuan) to cell_10 (cell2 of site Chunxiao), even the signal level of target cell fulfilled the condition to trigger A3 event, and there wasn’t any measured neighbour cell in the UE NCell info window. But from Figure 14.2-22, we found that UE received measConfig in source cell, which means the intra LTE mobility was enabled. To solve the issue, we did steps as below: Step1: Checked the ReportConfig in UeMeasurementConf of the source eNB, verified that there is no error in A3 event report configuration; Step2: Because UE could not measure any neighbour cell, so we checked the value of sMeasure (to allow UE to measure neighbour cell) parameter in source eNB, and found that the value is the maximum “-43”, which will always allow UE to measure neighbour cell, and no error exist. Step3: Checked the value of dlEARFCN in MeasObjectEUTRA of source eNB, and found that the dlEARFCN was set to 38050 which was different to the value 38150 used by the whole network. Step4: Modified the value of dlEARFCN to 37900 which are used now, and tested again to check the issue was solved.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 219/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-23: RSRP became better after optimization
UE measured neighbour cell Figure 14.2-24: source eNB received MR and UE successfully HO to target cell after optimization
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 220/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Measured carrierFreq was set to the right value Figure 14.2-25: Detail of measConfig message received in source cell after optimization After the optimization, we found that source eNB could receive UE reported measurement report as Figure 14.2-24, and cell_29 could successfully handover to cell_10, the RSRP in the test route became better too.
14.2.2.14.3
NEIGHBOUR CELL RELATIONSHIP MISSING ISSUE
We met this issue at site Pujiancai and Zhuzong in CMCC LST Zhangjiang area according to Figure 14.2-26.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 221/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 14.2-26: Test route and RSRP before optimization
UE reported many MR, but no HO
Figure 14.2-27: UE Signalling and Events in source cell before optimization
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 222/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Target cell was not configured in measConfig of source cell
UE reported measured result of target cell
Figure 14.2-28: Detail signalling message of measConfig and received measurement report in source cell before optimization From Figure 14.2-27, we found that UE reported many measurement reports which indicated target cell fulfilled event A3, but no handover occurred when UE went from cell_101 (cell3 of site Pujiancai) to cell_125 (cell3 of site Zhuzong). But from Figure 14.2-28, we found that target cell was not configured in UE received measConfig in source cell, which lead to the issue. To solve the issue, we did steps as below: Step1: Because there are few users in CMCC LST and no load when implement the test, RAC issue could be excluded; Step2: Checked that the target cell was not in the HO restrict list, this reason could be excluded; Step3: Checked that the target cell was not barred, this reason could be excluded too; Step4 : Checked the LteNeighboringCellRelation configuration of both eNBs, we found that there were no bi-direction neighbouring cell relation between the two cells, which lead to the issue. Step5: Added direction neighbouring cell relation in the two and tested again to check the issue was solved.
14.2.2.14.4
HOW TO AVOID PING PONG HO ISSUE
In some certain scenarios, UE may experience ping-pong handover between two badly-overlapped cells. The major reason is that two cells have a big overlapped area in which two cells alternate to emanate the stronger signal. As such, UE will switch between those two cells rapidly according to the RSRP when the parameters are not well tuned. To avoid ping pong effect between different cells ID’s as show in the Error! Reference source not found., a balance between the following arameters should be done: filterCoefficientRSRP - The higher the value of filterCoefficientRSRP the smoother the reported measurement will be and consequently the less likely ping-ponging occurs between sectors during handover Hysteresis - Increasing the value of this parameter would delay the HO due to the more important difference that must exist between the serving cell and neighbour cell. CellIdividualoffset – Low values decrease the number of ping pongs between the 2 cells and delay the HO. timeToTrigger – Increasing the value would delay the HO. eventA3Offset – Increasing the value would delay the HO. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 223/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
These parameters if setup correctly will help to avoid ping pong effect and improve throughput, during HO attempts.
HO Ping Pong area. Cell PCI 121 and 108
Figure 14.2-29: HO ping pong area between cell_121 and cell_108 (CMCC LST-Density urban)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 224/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
15 IRAT MOBLILTY OPTIMIZATION HINTS
Mobility
Idle Mode
UMTS Reselection
Active Mode
GERAN Reselection
UMTS PS Handover Release and Redirect Blind
GERAN Cell Change Order with NACC Release and Redirect Blind
15.1 INTER-RAT MOBILITY STRATEGY Hereunder is a example of the strategy of RAT selection when multimode UE switch on and select a RAT to camp on.
Figure 15.1-1: 2G/3G/LTE RAT selection example when multimode UE switch on Strategy of mobility in RRC Idle Mode (i.e. IRAT cell reseltction) and in RRC Connected Mode (redirection, PS HO, CSFB or CCO) between LTE and other RAT (Utran or Geran) should follow the requirements of the service venders and traffic QoS. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 225/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
LTE has wide band and can provide higher speed service than Utran-TDD (TD-SCDMA), but LTE uses higher frequency (e.g. band38) than Utran-TDD (e.g. band34) and gets poorer coverage when co-site, as example below:
Figure 15.1-2: LTE TDD/Utran TDD RAT coverage gap example So, the IRAT mobility strategy should consider the trade off between coverage and traffic performance (QoS). Hereunder is the strategy used in CMCC LST IRAT cell reselection between LTE TDD and Utran TDD test: Step1: Set priority of LTE cell to 7 (the highest priority), TD-SCDMA cell to 5 (lower priority than LTE). Step2: Tuning LTE IRAT cell reselection related parameters, i.e. the value of sNonIntraSearch, threshServingLow, threshXLow and tReselectionUtra. The aim is to keep UE stay in LTE TDD RAT as long as possible until it fulfill the cell reselection condition to Utran TDD RAT; Step3: Tuning Utran TDD IRAT cell reselection related parameters, i.e. thresh_priorit_search, thresh-Serving-low, ThreshX-High, and TreselectionUTRA. The aim is to let UE reselect to LTE TDD RAT as soon as possible when LTE RAT fulfill the cell camp on condition.
15.2 LTE-UMTS OPTIMIZATION HINTS Mobility from LTE to UTRA TDD has been implemented in three forms: Cell reselection: for mobility in idle mode PS Handover: for mobility in connected mode Release & Redirect: for mobility in connected mode Cell reselection from EUTRAN to UTRAN includes the support of additional information elements of SIB3 and SIB6 by the eNB. Redirection (including RRC connection release) includes the support of configuration of UE measurement and RRCConnectionRelease message with the IE redirectedCarrierInfo.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 226/290
LTE Optimization Handbook TLA6.0
EUTRA-to-UTRA PS Handover CELL_DCH UTRA-to-EUTRA PS Handover
MGR/TIPS/NEA
GSM_Connected E-UTRA RRC_CONNECTED
CELL_FACH
UTRA to E-UTRA Reselection
CCO, Reselection Connection establishment/release
RRC Connection establishment/Release
Connection establishment/release UTRA_Idle
GPRS Packet transfer mode
CCO with optional NACC, Redirection
Redirection CELL_PCH URA_PCH
Handover
UTRA to E-UTRA Reselection
E-UTRA RRC_IDLE
E-UTRA to UTRA Reselection
Reselection
GSM_Idle/GPRS Packet_Idle
CCO, Reselection
Figure 15.2-1: LTE to UTRAN mobility in the context of IRAT mobility In TLA6.0 and for the UE in RRC idle mode, the inter-RAT mobility to UTRA-TDD is supported with the cell reselection from E-UTRA-TDD to UTRA-TDD. Cell reselection inter-RAT (E-UTRA-TDD to UTRA-TDD) is internal to the UE and controlled by system information parameters provided in SystemInformationBlockType6 SIB6 and SystemInformationBlockType3 SIB3. Cell reselection to UTRA-TDD is supported with SystemInformationBlockType6 SIB6. Cell reselection to UTRA-TDD is enhanced with SystemInformationBlockType3 SIB3 (IE speedStateReselectionPars). In TLA6.0 and for the UE in RRC connected mode, the inter-RAT mobility to UTRA-TDD is supported with the RRC connection release and redirection from E-UTRA-TDD to UTRA-TDD. The redirection is driven by the eNodeB based on radio criteria. When the EUTRA serving cell becomes worse than a threshold and the UTRA-TDDneighbouring cell becomes better than another threshold, the eNodeB receives a Measurement Report message with Event-B2 from the mobile. If the UE capability or the network cannot support EUTRA-to-UTRA-TDD PS handover, the Algorithm for Control Procedures for Mobility decides to trigger a redirection EUTRA-to-UTRA-TDD. When the eNodeB does not receive any Measurement Report message from the mobile and, if the UE capability or the network cannot support EUTRA-to-UTRA-TDD PS handover, the selection of mobility mechanism decides to trigger a blind EUTRA-to-UTRA-TDD redirection (i.e. without measurements). The function of EUTRA-to-UTRA-TDD redirection, the eNodeB provides the following functions; (1) EUTRA-to-UTRA-TDD redirection execution phase; (2) EUTRA-to-UTRA-TDD redirection completion phase. During the previous phase of section Mobility Trigger Management (Control Procedures for Mobility), the source ENB has decided to initiate a EUTRA-to-UTRA-TDD redirection to the target access network (UTRA-TDD). The source ENB will give a command to the UE to re-select a cell in the target access network via the RRC CONNECTION RELEASE. The RRCConnectionRelease message is used to command the release of an RRC connection.
15.2.1 INTER- RAT MOBILITY IN RRC IDLE MODE Done by the UE under control from EUTRAN via System Information Broadcast Cell selection: the UE seeks to identify a suitable cell i.e. cell for which the measured cell attributes satisfy the cell selection criteria; if found it camps on that cell and starts the cell reselection procedure Cell reselection: UE performs measurements of the serving and neighbour cells: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 227/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Intra-frequency reselection is based on ranking of cells; Inter-frequency and Inter-RAT reselection is based on absolute priorities where UE tries to camp on highest priority frequency available. The cell selection and reselection algorithms are controlled by setting of parameters (thresholds and hysteresis values) that define the best cell and/or determine when the UE should select a new cell. SIB6 contains information about UTRA frequencies and UTRA neighbouring cells relevant for cell reselection (including cell re-selection parameters common for a frequency as well as cell specific reselection parameters) In RRC_IDLE mode, the cell reselection is internal to UE and is controlled by the System Information Parameters provided in SIB6 if the reselection to UTRA FDD is enabled (isMobilityToUtraAllowed = TRUE). Any modification of SIB6 parameters triggers a dynamic system information modification procedure
Figure 15.2-2: Cell Reselection procedure In order to limit the amount of inter-RAT measurements an additional criterion broadcasted in SIB3 is used: Snonintrasearch: threshold for serving cell reception under which the UE may trigger inter-RAT measurements for cell reselection. Configurable under : CellSelectionReselectionConf::sNonIntraSearch The UE applies rules as follows, where CRP = Cell Reselection Priority and IRAT=UTRAN, GERAN, S = selection criterion:
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 228/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-3: UE rules follow-up 3GPP R8 rules: SservingCell > 0 where SservingCell = Qrxlevmeas – (Qrxlevmin + Qrxlevminoffset) - Pcompensation * Pcompensation = compensation factor to penalize the low power UEs = 0 Let‘s consider: IF SServingCell > Snonintrasearch -> UE choose to not perform inter-RAT measurements If SServingCell ≤ Snonintrasearch -> UE shall perform inter-RAT measurements ... Now using parameters... IF Qrxlevmeas > Qrxlevmin + Qrxlevminoffset + Snonintrasearch -> UE does not measures IF Qrxlevmeas ≤ Qrxlevmin + Qrxlevminoffset + Snonintrasearch -> UE measures
UE will reselect the new cell if the conditions below are met: • Sservingcell < threshServingLow and SnonServingCell > threshXLow during tReselectionUtra • No cell with higher priority than the serving will fulfil the condition: SnonServingCell > threshXHigh during tReselectionUtra • More than 1 second(s) has elapsed since the UE camped on the current serving cell.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 229/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-4: LTE to UTRAN Mobility – (RSRP vs. Time) measurement phase Several database parameters are used for handling this kind of mobility. Some of them are tuneable and most important are the following: qRxLevMin, pMaxUTRA, qQualMin, sNonIntraSearch, trhreshServingLow, threshXLow. A summary of the selection procedure is presented in the figure below.
RSRP LTE Serving Cell
UE starts IRAT freq meas for lower cell priority reselection
4
1 Qrxlevmin(SIB3)+ sNonIntraSearch
UTRAN cell is reselected and Squal >0 CPICH or PCCPCH RSCP 2
tReselectionUTRA 3
Qrxlevmin(SIB3)+ threshServingLow
Lower priority UTRAN Cell LTE higher priority nonserving cell
2 Criteria 1 not met
Qrxlevmin+Qrxlevminoffs et +Pcompensation +threshXHigh Qrxlevmin+Qrxlevminoffs et +Pcompensation +threshXLow 3 time
Qrxlevmin, sNonIntraSearch, threshServingLow – SIB3 Qrxlevmin, Pcompensation, threshXHigh, threshXLow, tReselectionUTRA – SIB6(UTRAN) Qrxlevmin, Pcompensation, threshXHigh, threshXLow, tReselectionUTRA – SIB6 or SIB7
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 230/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-5: LTE to UTRAN Mobility – Algorithm Cell Reselection toward lower priority UTRAN Cell
Figure 15.2-6: LTE to UTRAN Mobility – (RSRP vs. Time) Decision Phase
15.2.1.1
QRXLEVMIN
Clarifications regarding qRxLevMin: A parameter with the same name appears in several objects and is then transmitted to UE inside several system information block types i.e. Sibs: CellSelectionReselectionConf – transmitted in SIB1 and SIB3 CellReselectionConfUtraFdd – transmitted in SIB6 CellReselectionConfUtraTdd – transmitted in SIB6 CellReselectionConfGERAN – transmitted in SIB7 The LTE – UTRAN mobility is using two of them, the one sent in SIB3 and the one sent in SIB6. The IE SystemInformationBlockType3 contains cell re-selection information common for intrafrequency, inter-frequency and/or inter-RAT cell re-selection (i.e. applicable for more than one type of cell re-selection but not necessarily all) as well as intra-frequency cell re-selection information other than neighbouring cell related. The IE SystemInformationBlockType6 contains information relevant only for inter- RAT cell reselection i.e. information about UTRAN frequencies relevant for cell re-selection. This parameter configures the minimum required RSCP level in the UTRAN cell, used by the UE in cell reselection. Recommended & Default Value= "-115"
NEA Recommended Value= This value depends on the strategy of the operator and the network coverage. In IRAT test (TD-LTE and TD-SCDMA) of CMCC LST phase 2, this value is 98dBm.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 231/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Note: The qRxLevMin should be cross-checked and tuned with threshXHigh or threshXLow according to cellReselectionPriority setting in CellReselectionConfUtraTdd/Fdd. Expected behaviour when changing this parameter Increasing the value of this parameter could: Determine a delayed selection of UTRAN cell, i.e. a shrinking of the UTRAN cell in idle mode. Decreasing the value of this parameter would: Determine an early selection of UTRAN cell which is similar to a shrinking of the EUTRAN cell. KPI Impact: Mobility - high values might create coverage discontinuity in idle, as seen by mobile.
Optimization of this parameter should contain the following steps Step 1: Set the value of qRxLevMin to one of the following values {- 124,-122,-120,-118,-116}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qRxLevMin and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.2
SNONINTRASEARCH
This parameter is used for setting a threshold for the selection criterion, threshold that would determine when, based in serving cell field level, the UE starts performing measurements for interfrequency and inter-RAT measurements. It is used for cell reselection. Recommended & Default Value= "16" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start earlier the measurement for inter-RAT reselection which will probably empty the UE battery sooner. Decreasing the value of this parameter would: Determine the UE to start later the measurements for inter-RAT reselection. Possible impact correct and timely reselection for high speed UEs. KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection. The optimization process should contain the following steps: Step 1: Set the value of sNonIntraSearch to one of the following values {12, 14, 16, 18, and 20}.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 232/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another sNonIntraSearch and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.3
THRESHSERVINGLOW
This threshold is used when the mobility towards lower priority frequency is taken in consideration. The default priority for UTRAN frequency is lower than for EUTRAN frequency which implies that this parameter is used each time mobility towards UTRAN happens. This parameter sets the threshold of the selection criteria in case of mobility towards lower priority RAT. The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionRAT. The parameter discussed here only impacts the part related to the serving cell. The condition on the serving cell can be rewritten as a condition on the measured level in the serving cell as follows:
Q relevmeas qRxLevMin threshServ ingLow The optimization of threshServingLow is based on this relation. Recommended & Default Value= "0"
NEA Recommended Value= This value depends on the strategy of the operator and the network coverage. In IRAT test (TD-LTE and TD-SCDMA) of CMCC LST phase 2, this value is 20 to make UE reselection easier to TD-SCDMA. Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of UTRAN cell, i.e. a shrinking of the EUTRAN cell in idle mode. Indeed, it is possible that modifying the value of this parameter in a given range does not in fact impact the selection due to possibly stronger condition on the UTRAN cell. Decreasing the value of this parameter would: Determine a later selection of UTRAN cell which is similar to a shrinking of the UTRAN cell. The similar observation made above, regarding the condition that ultimately triggers the selection is applicable for this situation as well. KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection. Coverage - high values might create coverage discontinuity during reselection operation. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 233/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Optimization of this parameter, in conjunction with threshXLow should aim at obtaining the cell sizes for both UTRAN cell and EUTRAN cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshServingLow should be decoupled from the optimization for threshXLow. For this, the value of threshXLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshServingLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshServingLow to one of the following values {0, 6, 12, 18, and 24}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshServingLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value.
15.2.1.4
THRESHXLOW
This threshold is used when the mobility towards lower priority frequency is taken in consideration. The default priority for UTRAN frequency is lower than for EUTRAN frequency which implies that this parameter is used each time mobility towards UTRAN happens. This parameter sets the threshold of the selection criteria in case of mobility towards lower priority RAT. The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionRAT. The parameter discussed here only impacts the part related to the neighbouring cell. The condition on the neighbouring cell can be rewritten as a condition on the measured level in the serving cell as follows: Qrelevmeas > Qrxlevmin + Pcompensation + threshXLow The optimization of threshXLow is based on this relation. Recommended & Default Value= "0"
NEA Recommended Value: This value depends on the strategy of the operator and the network coverage. In IRAT test (TD-LTE and TD-SCDMA) of CMCC LST phase 2, this value is 18.
Expected behaviour when changing this parameter Decreasing the value of this parameter could: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 234/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Determine an earlier selection of UTRAN cell, i.e. a shrinking of the EUTRAN cell in idle mode. Indeed, it is possible that modifying the value of this parameter in a given range does not in fact impact the selection due to possibly stronger condition on the EUTRAN cell. Increasing the value of this parameter would: Determine a later selection of UTRAN cell which is similar to a shrinking of the UTRAN cell. The similar observation made above, regarding the condition that ultimately triggers the selection is applicable for this situation as well. KPI Impact: Mobility - high values might create coverage discontinuity during reselection operation due to shrinking UMTS cell as seen by the UE. Optimization of this parameter, in conjunction with threshServingLow should aim at obtaining the cell sizes for both UTRAN cell and EUTRAN cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshXLow should be decoupled from the optimization for threshServingLow. For this, the value of threshServingLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshXLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshXLow to one of the following values {0, 6, 12, 18, and 24}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshXLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate. Step 5: Choose the optimized value.
15.2.1.5
TRESELECTIONUTRA
This parameter concerns the cell reselection timer tReselectionRAT for UTRAN. Broadcast in SystemInformationBlockType6. It imposes a condition on the reselection. UE will actually reselect the new cell, only if the new cell is better ranked than the serving cell during a time interval tReselectionUtra. Recommended & Default Value= "2" Expected behaviour when changing this parameter Increasing the value of this parameter could: Determine a delayed reselection which could be an issue for fast moving UEs. Decreasing the value of this parameter would: Facilitate ping-pong behaviour during reselection process.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 235/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Optimization of this parameter should find a trade-off between delayed reselection and ping pong behaviour. Most probably, if the UEs are not moving fast, the delayed reselection would not be an issue. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionUtra to one of the following values {1, 2, 3, and 4}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionUtra. Step 4: Post-process the logs and analyze them as reselection position vs. tReselectionUtra values and ping pong behaviour vs. tReselectionUtra values and choose the optimized value to obtain smallest interruption time and highest success rate. Step 5: Calculate the cell reselection success rate in each direction.
15.2.1.6
TRESELECTIONUTRASFMEDIUM
This parameter contributes to the configuration of the IE SystemInformationBlockType6 if the UE is in Medium Mobility state. The concerned mobility control related parameter is multiplied with this factor if the UE is in Medium Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended & Default Value= “0dot5” Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier.
KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later.
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionUtraSfMedium to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 236/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 3: Repeat Step 2 for another value of tReselectionUtraSfMedium. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.7
TRESELECTIONUTRASFHIGH
This parameter contributes to the configuration of the IE SystemInformationBlockType6 if the UE is in High Mobility state. The concerned mobility control related parameter is multiplied with this factor if the UE is in High Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended & Default Value= "oDot25" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later.
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionUtraSfHigh to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionUtraSfHigh. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.8
NCELLCHANGEHIGH
This parameter configures the number of cell changes to enter high mobility state Recommended & Default Value= "12" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 237/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later.
A procedure that optimizes nCellChangeHigh would contain the following steps: Step 1: Set the value of nCellChangeHigh to one of the following values {10, 11, 12, 13, and 14}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.9
NCELLCHANGEMEDIUM
This parameter configures the number of cell changes to enter medium mobility state Recommended & Default Value= "4" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later.
A procedure that optimizes nCellChangeMedium would contain the following steps: Step 1: Set the value of nCellChangeMedium to one of the following values {1, 2, 3, 4, 5, and 6}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.10 QHYSTSFHIGH This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-High included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in High Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 238/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later.
A procedure that optimizes qHystSfHigh would contain the following steps: Step 1: Set the value of qHystSfHigh to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.1.11 QHYSTSFMEDIUM This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-Medium included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in Medium Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 239/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
A procedure that optimizes qHystSfMedium would contain the following steps: Step 1: Set the value of qHystSfMedium to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.2 FIELD TEST RESULTS OF INTER- RAT MOBILITY IN RRC IDLE MODE In CMCC LST, the IRAT cell reselection test was implemented in Zhangjiang area, as the test route below:
Figure 15.2-7: LTE TDD/Utran TDD IRAT cell reselection driver test route We used 2 dual-modes UEs to implement driver test around the test route, need at least 30 times from LTE TDD to Utran TDD and vice versa.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 240/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-8: Detail signalling and events example of LTE TDD/Utran TDD IRAT cell reselection Based on the Figure 15.2-8, we can find that UE will initial RAU procedure while cell reselect from LTE TDD to Utran TDD, UE will initial TAU procedure while cell reselect from Utran TDD to LTE TDD.
Hereunder are the LTE TDD<-->Utran TDD IRAT cell reselection test results in CMCC LST: IRAT cell reselection test results UE1 TD-L-->TD-S TD-S-->TD-L
Cell reselection attempt times 37 36
Cell reselection success rate 94.595% 97.222%
Avg. cell reselection latency(s) 4.549 2.411
Signal level before cell reselection -105.740 -74.271
Signal level after cell reselection -72.940 -90.958
IRAT cell reselection test results UE2 TD-L-->TD-S
Cell reselection attempt times 35
Cell reselection success rate 91.429%
Avg. cell reselection latency(s) 4.448
Signal level before cell reselection -107.208
Signal level after cell reselection -71.917
TD-S-->TD-L 32 100.000% 3.432 -71.130 -88.783 Table 15-1: IRAT cell reselection between LTE TDD & Utran TDD test results in CMCC LST (Density Urban) Based on the test results, we can find that different UE may get different cell reselection success rate in the whole test route, this may because of the difference between UEs, and the avg. cell reselection latency of LTE TDD to Utran TDD is higher than Utran TDD to LTE TDD, this may caused by Ericsson 3G GGSN configuration which we cannot manage. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 241/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
From the latency distribution of the test results, we can find that RAU procedure or TAU procedure occupy the most part of the whole latency. The software used in LTE TDD platform during IRAT test is as below: CMCC LST SW versions HSS
R3.0 IAP1
PCRF-1
DSC_4_0_R3
S-GW
R3.1_R3
P-GW
R3.1_R3
MME02
R25.55.20 ENB_TA0400_D00_E00027 + database MIM 11.4.1 ed06 + NEM.LA4.0.1_D1.9 + WPS_MIM_11.4.1_ed06 ZTE-A370 (dual-mode UE), sw: LTMV1.0.2B01P02, PC drive sw: PCW_ZTEMF820S2V1.0.0B02
eNB40 UE CDS
V7.0_17B120422
EPS Integrity Protection Algorithm EPS Encryption Algorithm Table 15-2: LTE TDD SW Reference MME02
128-EIA0 (No Algorithm) 128-EEA0 (No ciphering)
Note: Only LTE TDD RAN & ePC side is ALU product, Utran TDD RAN & RNC side is Huawei product, 3G GGSN/SGSN is Ericsson product, we can only provide software info on ALU product.
15.2.3 INTER-RAT MOBILITY IN RRC CONNECTED MODE The inter-RAT mobility in RRC_CONNECTED state is UE assisted (since the UE provides measurements) and consists in a handover (HO) controlled by the network, with a HO preparation signalling in EUTRAN and UTRAN. Typically, the UE measurement reporting to eNB triggers the handover preparation. In addition, the handover decisions may take other inputs, such as neighbour cell load (not LA3.0), Traffic distribution (eMCTA), transport and hardware resources (not LA3.0) and Operator defined policies into account. In TLA6.0, the inter-RAT measurements on the UTRA TDD overlay are UTRA TDD PCCPCH RSCP, UTRA TDD carrier RSSI, UTRA TDD PCCPCH Ec/No. The reporting of UE measurements is event-triggered and configured in the UE by the EUTRAN eNodeB.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 242/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-9: Events state machine
15.2.3.1
UE MEASUREMENTS NEEDED FOR PS HO TO UTRA-TDD
Figure 15.2-10: UE measurements needed for PS HO to UTRA-TDD Intra-freq measurements to trigger inter-RAT measurements The eNodeB configures an event A2 (A2_CA for Coverage Alarm) that is configured at cell entry (call set-up, incoming handover, RRC re-establishment) with purpose in MiM set to “Entering-CoverageAlarm” (serving worse than mobility threshold) Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 243/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Inter-RAT measurements for PS HO to UTRA The eNodeB configures an event B2 (Serving worse than Threshold1 and neighbour better than Threshold2) that is configured after reception of an event A2_CA i.e. when the radio enter the coverage alarm conditions An event B1 (Neighbour better than Threshold) can also be configured for CS fallback as explained by TLA6.0 CSFB feature presentation Measurement Gaps may be needed with respect to UE capabilities (per RAT and carrier) The eNodeB checks conditions before configuring inter-RAT measurements to UTRA-TDD the mobility to UTRAN is activated in MiM (isMobilityToUtranAllowed ‘TRUE’) although PS HO may be deactivated (isPsHoToUtranAllowed ‘FALSE’) since redirection may be used At least one inter-RAT neighbour carrier is configured in MiM for the serving LTE cell UE can perform inter-RAT measurements, reporting and measurement reporting event B2 in eUTRA RRC_connected
15.2.3.1.1 PS HO PREPARATION
Figure 15.2-11: Call flow for PS HO – Preparation phase RRC Measurement reporting (event B2) The eNodeB receives a RRC MeasReport with event B2 with a Measurement Purpose set to “MobilityInter-RAT-to-UTRA” as retrieved in the call context: Measurement Purpose retrieved from the MeasId in RRC MeasReport Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 244/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The eNodeB takes a “PS-HO-to-UMTS” decision with data configured in MiM If PS handover is to be performed, eNB will select the best UTRA cell reported by the UE as the target cell that supports the serving PLMN (to favor intra-PLMN handover). If there is no intra-PLMN target cell available, eNB will select the best UTRA cell as the target cell that supports one of the equivalent PLMNs. The eNB retrieves the HO target Cell/RNC from the MeasObject stored in the eNB call context The UTRAN carrier reported by the UE leads to the UtraTddNeighboringFreqConf object (MiM) The UTRAN PhysicalCellId (primary scrambling code) leads to the UtraTddNeighboringCellRelation (MiM) The selected target cell should not in the forbidden LACs included in the HandoverRestrictionList for the call and the controlling RNC of the cell should have RncAccess::psHandoverUtraEnabled set to ‘True’. The target cell selection for emergency CSFB alls is not restricted by the HandoverRestrictionList. the eNB retrieves the target RNC and the DL forwarding tunnel type (direct or indirect) the RNC state is given by the UtraTddNeighboringCellRelation.RncAccess the user-plane tunnel type by the RncAccess.DirectFwdPathAvailability of course the UE must support the PS HO to 3G as reported by “FGI bit#8 - EUTRA RRC_CONNECTED to UTRA CELL_DCH PS handover” The S1AP HO Preparation Procedure is triggered by source eNB towards target RNC There are exchanges of RRC containers that are transparent to the eNodeB (mediation service) eNB will send a Handover Required message to the MME and start timer TS1relocprep with duration PsHoToUtraTimersConf::tS1RelocPrepForPsHandoverToUtra (the PsHoToUtraTimersConf instance is pointed to by the RncAccess::psHoToUrtaTimersConfId associated with the selected UtraTddNeighboringCellRelation). The SourceToTarget container in S1AP HANDOVER REQUIRED with a source-RNC to target-RNC radio container and UE UTRAN capabilities that are sent from UE to the target RNC. This information is received from the UE (UE capabilities enquiry, RRC procedure) or from the MME the TargetToSource container in S1AP HANDOVER COMMAND with a target-RNC to source-RNC radio container and UTRAN access info about the target UTRA cell sent from the target RNC to the UE
15.2.3.1.2 PS HO EXECUTION
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 245/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-12: Call flow for PS HO – Execution phase
Once a Handover Command message is received from MME, eNB will stop timer TS1relocprep and start timer TS1relocoverall with duration PsHoToUtraTimersConf::tS1RelocOverallForPsHandoverToUtra (the PsHoToUtraTimersConf instance is pointed to by the RncAccess::psHoToUrtaTimerConfId associated with the selected UtraTddNeighboringCellRelation). eNB will send a MobilityFromEutraCommand to the UE with purpose set to ‘handover’ and targetRAT-Type set to ‘utra’. If ActivationService::isUtraDataForwardingAllowed is set to ‘True’, eNB will start data forwarding for each E-RAB listed in E-RABSubjecttoDataForwardingList received in the Handover Command message from MME. RRC MobilityFromEUTRACommand This message includes the transmission of the radio container that is the target-RNC to source-RNC radio container previously received in the S1AP HANDOVER COMMAND message. Indeed the eNodeB has ensured the S1AP-to-RRC mediation in a transparent fashion.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 246/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-13: PS HO to UTRA-TDD - End-to-End call flows
Figure 15.2-14: LTE to UTRAN Mobility – Redirection Execution
15.2.3.2
REDIRECTION TO UTRAN
Inter-RAT Mobility to UMTS in RRC Connected Mode: RRC Connection Release and Redirection to UTRAN. Redirection from LTE to a UTRAN target cell relies on radio measurements to trigger the redirection procedure. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 247/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
A redirection results in a RRC Connection Release from the source eNB, instructing the UE to leave the LTE eUTRAN and start access on a new target cell in the UTRAN RAT. Only blind redirection was implemented in TLA6.0. Blind redirection means redirection without measurements on a target RAT. Blind redirection is triggered by detection of serving cell degradation (eventA2) when intra-frequency LTE radio conditions fall below a configured threshold. Blind Redirection towards another RAT (e.g. UMTS) A2_floor_threshold in the diagram below is thresholdEutraRsrp or thresholdEutraRsrq.
1. Serving radio level goes below the A2_floor_threshold. 2. timeToTrigger expires (thick red line) 3. Meas Report Event A2 with purpose ‘Blind-Redirection-To-3GPP-RAT’ sent to eNB 4. eNB performs control procedure for Blind Redirection to UTRAN.
Figure 15.2-15: RAT frequency with highest cellReselectionPriority is chosen for redirection
In RRC_Connected mode, the eNB may trigger a procedure of RRC Release with Redirection Information (IE redirectedCarrierInfo) - if this is enabled if isMobilityToUtraAllowed=TRUE - and if the UE is also be eligible for redirection to UTRA-TDD (support UTRA-TDD ) - and if mobility to UTRA not forbidden for UE in S1AP HandoverRestrictionlist. The redirection can be blind (eMCTA used for redirectedCarrierInfo) or based on inter RAT measurement to UTRA-TDD . Redirection to UTRAN triggers: A MeasReport with event B2 and purpose “Mobility inter-RAT to UTRA”: this is the Meas- Based Redirection when the PS HO cannot be performed A MeasReport with event A2 and purpose “Below Serving Floor”: Redirection Blind (MobilityPriorityTable::defaultConnectedPriorityOfFreq by eMCTA for blind) or Meas-Based (measured UTRAN carrier, if PS HO was ongoing)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 248/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-16: RRC Connection Release with Redirection Info from EUTRAN to UTRAN Event B2 – Serving becomes worse than threshold1 and inter-RAT neighbour becomes better than threshold2. Entering conditions for this event:
& Ms = measurement result of the serving cell [dBm] Hys = reportConfigUTRA::hysteresis [dB] Thresh1 = ReportConfigUTRA::thresholdEutraRsrpB2 [dBm] Mn = measurement result of the inter-RAT neighbour cell [dBm] Ofn = MeasObjectUTRA::offsetFreq, corresponding to the neighbouring cell [dB] Thresh2 = ReportConfigUTRA::thresholdUtraRscp [dBm]
Figure 15.2-17: Inter RAT threshold for event B2 Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 249/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
UE
Source ENB
Source MME
Triggers: - an A2 measurement report is received - a B2 measurement report is received - a CS Fallback is triggered
Redirection and Release Initiation RRC CONNECTION RELEASE releaseCause::=other redirectedCarrierInfo::=utra-FDD or utra-TDD > ARFCN-ValueUTRA (optional) cellInfoListUTRA-FDD-r9
UE CONTEXT RELEASE REQUEST MME-UE-S1AP-ID ENB-UE-S1AP-ID Cause=interrat-redirection
The UE selects a suitable cell on the UTRAN frequency indicated by the RedirectedCarrierInfo
Release Completion ENB releases the UE context and associated resources
UE CONTEXT RELEASE COMMAND MME-UE-S1AP-ID ENB-UE-S1AP-ID Cause=normal-release
UE CONTEXT RELEASE COMPLETE MME-UE-S1AP-ID ENB-UE-S1AP-ID MME keeps the UE context
MME releases associated S1 resources
UE
Source ENB
Source MME
Figure 15.2-18: Call flow for redirection from EUTRAN to UTRAN-Overview
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 250/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.2-19: Call flow for redirection from EUTRAN to UTRAN-Description
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 251/290
LTE Optimization Handbook TLA6.0
15.2.3.3
MGR/TIPS/NEA
THRESHOLDEUTRARSRPB2
This parameter sets the RSRP threshold for the serving cell of the selection criteria in case of mobility towards UTRAN. Recommended Value= "-100" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of UTRAN cell, i.e. a shrinking of the EUTRAN cell in active mode. Decreasing the value of this parameter would: Determine a later selection of UTRAN cell, which is similar with a shrinking of the UTRAN cell. KPI Impact: Mobility - low values of this parameter might create coverage discontinuity during selection operation due to shrinking UMTS cell as seen by the UE.
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of thresholdEutraRsrpB2 to one of the following values {-104,-102,-100,-98,-96}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Step 3: Repeat Step 2 for another value of thresholdEutraRsrpB2. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.4
THRESHOLDUTRARSCP
This parameter sets the RSRP threshold for the target cell of the selection criteria in case of mobility towards UTRAN.
Recommended Value= "-114" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a later selection of UTRAN cell, which is similar with a shrinking of the UTRAN cell in active mode. Decreasing the value of this parameter would: Determine an earlier selection of UTRAN cell, which is similar with a shrinking of the EUTRAN cell. KPI Impact: Mobility – High value of this parameter might create coverage discontinuity during selection operation due to shrinking UMTS cell as seen by the UE. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 252/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
For optimization, a procedure containing the following steps can be used: Step 1: Set the value of thresholdUtraRscp to one of the following values {-118,-116,-114,-112, -110}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Step 3: Repeat Step 2 for another value of thresholdUtraRscp. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.5
OFFSETFREQUTRA
This parameter is used to indicate a frequency specific offset to be applied when evaluating triggering conditions for measurement reporting. Recommended Value= "0" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of UTRAN cell, which is similar with a shrinking of the EUTRAN cell in active mode. Decreasing the value of this parameter would: Determine a later selection of UTRAN cell, which is similar with a shrinking of the UTRAN cell. KPI Impact: Mobility - low values of this parameter might allow the UE to determine the strongest cell later. High values of this parameter might allow the UE to determine the strongest cell earlier. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of offsetFreqUtra to one of the following values {-3,-2,-1, 0, 1, 2, 3}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Step 3: Repeat Step 2 for another value of offsetFreqUtra. Step 4: Post process the logged data and determine the positions at which the UE selected the UTRAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.6
FILTERCOEFFICIENTOFQUANTITYCONFIGUTRA
This parameter is used to configure the IE filterCoefficient of QuantityConfigUtra. The parameter is optional and is required only when inter-RAT mobility to UTRAN is supported. If this parameter is not configured (absent) then the default RRC value defined in 36.331 is used by the eNB and signalled to the UE. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 253/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The RSRP values reported by the UE are obtained by filtering several measurements performed by the UE. If this filter can allow quick variation to be reported or it can rely more on the last reported value and less on the measured value such that there is less variation in the sequence of the reported value. The higher the value of filterCoefficientOfQuantityUtra the smoother the reported measurement will be and consequently the less likely ping-ponging occurs between sectors during handover. Recommended Value= "fc4" Expected behaviour when changing this parameter: Increasing the value of this parameter would: Decrease the variation in the reported RSRP value. Decrease ping-pong between the cells in case of handover conditions. Delay the speed at which the reported RSRP adapts to the RSRP variation. This might eventually slightly delay the HO, if the value of the parameter is too high. Improve the system behaviour regarding the throughput during HO. Decreasing the value of this parameter would: Increase the variation in the reported RSRP value due to noise. Increase the ping-pong between the cells in case of handover conditions due to variations in reported RSRP. Decrease the HO quality relative to throughput. Increase the speed at which the reported RSRP adapts to the RSRP variation. This might speed up the HO which could manifest as ping-pong. KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Optimization of this parameter should be performed in conjunction with optimization of hysteresis and timeToTrigger parameters. Finding the optimum pair of (filterCoefficientOfQuantityUtra, hysteresis, and timeToTrigger) should consider the following steps: Step 1: Set the values of filterCoefficientOfQuantityUtra and to hysteresis and to timeToTrigger to one of the following {(fc4, 4,100), (fc5, 5, 80), (fc3, 3,200), (fc6, 6, 40)}, in both current cell and neighbour cell. Step 2: Perform a drive test while performing a download and log the throughput values and the position of the UE. Drive in and out of the current cell to the neighbour cell. Step 3: Repeat Step 2 for another pair of values of the three tested parameters. Step 4: Represent throughput vs. position (distance) (Service continuity), #HO-attempts, Success Rate/Failure Rate, #of Ping-pongs, HO interruption time for all pairs of tested values. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.7
HYSTERESIS
This IE is a parameter used within the entry and leave condition of an event triggered reporting condition. This is used to provision IE Hysteresis in IE ReportConfigInterRAT, in IE MeasConfig . This parameter defines the hysteresis used by the UE to trigger an inter-RAT event-triggered measurement report. It is used in several processes: Event B2 (Serving becomes worse than threshold1 and inter RAT neighbour becomes better than threshold2); Event B1 (Inter RAT neighbour becomes better than Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 254/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
threshold); Event A1 (Serving becomes better than threshold); Event A2 (Serving becomes worse than threshold); Event A3 (Neighbour becomes offset better than serving); Event A4 (Neighbour becomes better than threshold); Event A5 (Serving becomes worse than threshold1 and neighbour becomes better than threshold2).
Recommended Value= "4" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO due to the more important difference that must exist between the serving cell and neighbour cell. Drop the call if the value is too large i.e. connection to the serving cell is lost before having reached the neighbour cell level that satisfies the HO condition. Decreasing the value of this parameter would: Create a ping – pong behaviour because the measurement quick variations (noise-like) might trigger HO decisions. KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Throughput - low values of this parameter can generate a ping pong behaviour which can result in interruption times and low throughput during HO operation.
Optimization of this parameter should be performed in conjunction with optimization of filterCoefficientOfQuantityUtra and timeToTrigger parameters, as presented in the previous paragraph.
15.2.3.8
TIMETOTRIGGER
This parameter sets the time duration time during which the conditions to trigger an event report have to be satisfied before sending a RRC measurement report in event triggered mode. Recommended Value= "ms100" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO decision. Determine a call drop due to significant serving cell signal degradation before timeToTrigger expires. Decreasing the value of this parameter would: Generate ping-pong HO behaviour due to the fact that quick variations of the measured signal (noise-like variations) might satisfy the HO relation for the short while represented by timeToTrigger but not much longer. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 255/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell.
This parameter should be carefully optimized, best in conjunction with filterCoefficientOfQuantityUtra and hysteresis as presented in paragraph 15.2.3.6 and 15.2.3.7. Indeed, the optimized value can be impacted by the load of the surrounding cells.
15.2.3.9
REPORTINTERVAL
This parameter configures the IE reportInterval included in the IE ReportConfigInterRAT in the MeasConfig IE. The ReportInterval indicates the interval between periodical reports. The ReportInterval is applicable if the UE performs periodical reporting (i.e. when reportAmount exceeds 1), for triggerType ‘event’ as well as for triggerType ‘periodical’. Recommended Value= "ms240" Expected behaviour when changing this parameter Increasing or decreasing the value of this parameter would: Will not decrease or increase the Handover Success rate; but no point to have high values. .
KPI Impact: Mobility – low values of this parameter will decrease the HO success rate. High values of this parameter will increase the Ho success rate. A procedure that optimizes reportInterval would contain the following steps: Step 1: Set the value of reportInterval to one of the following values {120, 240, 480, 640, 1024, and 2048}. Step 2: perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportInterval. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.10 MAXREPORTCELLS This parameter defines the maximum number of cells to be reported in a measurement report. Recommended Value= "1"
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 256/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to report as many new neighbour cells as possible in a short time. Decreasing the value of this parameter would: Determine the UE to report fewer neighbour cells. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells. A procedure that optimizes maxReportCells would contain the following steps: Step 1: Set the value of maxReportCells to one of the following values {1, 2, 3, 4, 5, 6, 7, and 8}. Step 2: perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of maxReportCells. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
15.2.3.11 REPORTAMOUNT This parameter configures the number of periodical reports the UE has to transmit after the event was triggered. Recommended Value= "r8" Expected behaviour when changing this parameter Increasing the value of this parameter would: Increase the Handover Success Rate for multiple repetitions in bad RF conditions. Decreasing the value of this parameter would: Decrease the Handover Success Rate for multiple repetitions in bad RF conditions. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells. A procedure that optimizes reportAmount would contain the following steps: Step 1: Set the value of reportAmount to one of the following values {r1, r2, r4, r8, r16, r32, r64}. Step 2: Perform a drive test back and forth between the EUTRAN cell and UTRAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportAmount. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 257/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
15.3 LTE-GSM MOBILITY OPTIMIZATION HINTS Mobility from LTE to GSM has been implemented in three forms: Cell reselection and redirection (blind or measurement based redirection) LTE to GERAN mobility capability for a dual-mode UE in both RRC idle and connected modes. Cell Change Order mobility procedure & Network Assisted Cell Change from EUTRAN to GERAN. This feature supports basic mobility for UE moving from LTE radio coverage to GSM radio coverage. The capability provided by this feature enables the LTE-to-GSM mobility of a dual-mode UE in RRC_IDLE mode, which allows a UE leaving LTE coverage to recover service in GSM coverage, as soon as it gets available, i.e. radio conditions are sufficiently good. It also enables LTE-to-GSM mobility in RRC_CONNECTED with packet data session that is the leaving of an LTE coverage island, while the user is moving this done via PS Handover procedure. The triggering condition is because of radio conditions on LTE being degraded. Release/Redirect mechanism is supported to accommodate the scenarios where the optimized HO procedures. CCO with or without Network Assisted Cell Change (NACC) mechanism is supported. CCO mechanism is supported to accommodate the scenarios where the optimized PS-HO procedure is not supported in the UE, or the ePC Core, or the GERAN network during early deployment.
15.3.1 IDLE MODE
For cell reselection the UE must be in RRC-IDLE mode and to be GERAN capable. It shall receive the information about GERAN coverage through SIB7 message. Then the UE applies inter-RAT cell reselection criteria.
Figure 15.3-1: Reselection from eUTRAN to GERAN
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 258/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.3-2: LTE to GERAN Mobility – HO to GERAN cell
3GPP rules: SservingCell > 0 where SservingCell = Qrxlevmeas – (Qrxlevmin + Qrxlevminoffset) - Pcompensation * Pcompensation = compensation factor to penalize the low power UEs = 0 Let‘s consider: IF SServingCell > Snonintrasearch -> UE choose to not perform inter-RAT measurements If SServingCell ≤ Snonintrasearch -> UE shall perform inter-RAT measurements ... Now using parameters ... IF Qrxlevmeas > Qrxlevmin + Qrxlevminoffset + Snonintrasearch -> UE does not measures IF Qrxlevmeas ≤ Qrxlevmin + Qrxlevminoffset + Snonintrasearch -> UE measures
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 259/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Measurement phase RSRP > Qrxlevmin + Qrxlevminoffset + Snonintrasearch ≤ Qrxlevmin + Qrxlevminoffset + Snonintrasearch Measurement phase
Figure 15.3-3: LTE to GERAN Mobility (RSRP vs. Time) – Cell Reselection – Measurement phase UE will reselect the new cell if the conditions below are met: Sservingcell < threshServingLow and SnonServingCell > threshXLow during tReselectionGeran No cell with higher priority than the serving will fulfil the condition: SnonServingCell > threshXHigh during tReselectionGeran More than 1 second(s) has elapsed since the UE camped on the current serving cell.
Figure 15.3-4: LTE to GERAN Mobility – Cell Reselection toward lower priority GERAN cell Step 1: Serving cell become less good and the RSRP level decrease under [Qrxlevmin(SIB3) + Qrxlevminoffset + Pcompensation + sNonIntraSearch]. Then Measurement GAP is activated and the UE can detect and measure lower priority cells than the serving. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 260/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 2: Serving cell becomes worse and the RSRP level decrease under [Qrxlevmin(SIB3) + Qrxlevminoffset + Pcompensation + threshServingLow]. Cell reselection would be possible, but not yet candidate cell, reaching [Qrxlevmin(SIB7)+Qrxlevminoffset +Pcompensation+threshXLow]. In this user case, 1 and 2 occur at the same time because we have chosen to implement sNonIntraSearch= threshServingLow. Step 3: The situation just above is still reached and also, in the target cell, threshold [Qrxlevmin(SIB7)+Qrxlevminoffset+Pcompensation+threshXLow] is reached. tReselectionGeran is started. During tReselectionGeran, NO higher cell priority reaches [Qrxlevmin(SIB3 or SIB6)+Qrxlevminoffset+Pcompensation+threshXHigh] Step 4: tReselectionGeran is achieved, reselection is triggered.
Figure 15.3-5: LTE to GERAN Mobility (RSRP vs. Time) – Cell Reselection – Decision phase
15.3.1.1
QRXLEVMIN
Clarifications regarding qRxLevMin: A parameter with this name appear in several objects and is then transmitted to UE inside several system information block types i.e. Sibs: CellSelectionReselectionConf – transmitted in SIB1 and SIB3 CellReselectionConfUtraFdd – transmitted in SIB6 CellReselectionConfUtraTdd – transmitted in SIB6 CellReselectionConfGERAN – transmitted in SIB7 The LTE – GERAN mobility is using two of them, the one sent in SIB3 and the one sent in SIB7. The IE SystemInformationBlockType3 contains cell re-selection information common for intrafrequency, inter-frequency and/or inter-RAT cell re-selection (i.e. applicable for more than one type of cell re-selection but not necessarily all) as well as intra-frequency cell re-selection information other than neighbouring cell related. The IE SystemInformationBlockType7 contains information relevant only for inter- RAT cell reselection i.e. information about GERAN frequencies relevant for cell re-selection. This parameter configures the minimum required RSRP level in the GERAN cell, used by the UE in cell reselection.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 261/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "-101" Note: The qRxLevMin should be cross-checked and tuned with threshXHigh or threshXLow according to cellReselectionPriority setting in CellReselectionConfGERAN. Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a delayed selection of GERAN cell, i.e. a shrinking of the GERAN cell in idle mode. Decreasing the value of this parameter would: Determine an early selection of GERAN cell which is similar to a shrinking of the EUTRAN cell. KPI Impact: Mobility - high values might create coverage discontinuity in idle, as seen by UE. The optimization process should contain the following steps: Step 1: Set the value of qRxLevMin to one of the following values {- 105, -103, -101, -99, -97}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qRxLevMin and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.2
SNONINTRASEARCH
This parameter is used for setting a threshold for the selection criterion, threshold that would determine when, based in serving cell field level, the UE starts performing measurements for interfrequency and inter-RAT measurements. It is used for cell reselection. Recommended & Default Value= "16" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start earlier the measurement for inter-RAT reselection which will probably empty the UE battery sooner. Decreasing the value of this parameter would: Determine the UE to start later the measurements for inter-RAT reselection. Possible impact correct and timely reselection for high speed UEs.
KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 262/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
The optimization process should contain the following steps: Step 1: Set the value of sNonIntraSearch to one of the following values {12, 14, 16, 18, and 20}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another sNonIntraSearch and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.3
THRESHSERVINGLOW
This threshold is used when the mobility towards lower priority frequency is taken in consideration. The default priority for GERAN frequency is lower than for EUTRAN frequency which implies that this parameter is used each time mobility towards GERAN happens. This parameter sets the threshold of the selection criteria in case of mobility towards lower priority RAT. The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionRAT. The parameter discussed here only impacts the part related to the serving cell. The condition on the serving cell can be rewritten as a condition on the measured level in the serving cell as follows:
Q relevmeas qRxLevMin threshServ ingLow The optimization of threshServingLow is based on this relation. Recommended & Default Value= "0"
NEA Recommended Value= This value depends on the strategy of the operator and the network coverage. In IRAT scenario, this value can be set higher to make UE reselection easier to Geran. Expected behaviour when changing this parameter Increasing the value of this parameter could: Determine an earlier selection of GERAN cell, i.e. a shrinking of the EUTRAN cell in idle mode. Indeed, it is possible that modifying the value of this parameter in a given range does not in fact impact the selection due to possibly stronger condition on the GERAN cell. Decreasing the value of this parameter would: Determine a later selection of GERAN cell which is similar to a shrinking of the GERAN cell. The similar observation made above, regarding the condition that ultimately triggers the selection is applicable for this situation as well.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 263/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values delay the start of measurements performed by the UE which can be reflected in delayed reselection. Coverage – high values might create coverage discontinuity during reselection operation.
Optimization of this parameter, in conjunction with threshXLow should aim at obtaining the cell sizes for both GERAN cell and EUTRAN cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshServingLow should be decoupled from the optimization for threshXLow. For this, the value of threshXLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshServingLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshServingLow to one of the following values {0, 6, 12, 18, and 24}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshServingLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.4
THRESHXLOW
This threshold is used when the mobility towards lower priority frequency is taken in consideration. The default priority for GERAN frequency is lower than for EUTRAN frequency which implies that this parameter is used each time mobility towards GERAN happens. This parameter sets the threshold of the selection criteria in case of mobility towards lower priority RAT. The reselection criterion is quite a complex one which means that the optimization of this parameter would need some decoupling to be performed and the optimization to be made one parameter at a time. There is a condition on the serving cell through threshServingLow, another one on target cell through threshXLow and another one on time through tReselectionRAT. The parameter discussed here only impacts the part related to the neighbouring cell. The condition on the neighbouring cell can be rewritten as a condition on the measured level in the serving cell as follows:
Qrelevmeas > Qrxlevmin + Pcompensation + threshXLow The optimization of threshXLow is based on this relation. Recommended & Default Value= "0"
NEA Recommended Value: This value depends on the strategy of the operator and the network coverage. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 264/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Expected behaviour when changing this parameter Decreasing the value of this parameter would: Determine an earlier selection of GERAN cell, i.e. a shrinking of the EUTRAN cell in idle mode. Indeed, it is possible that modifying the value of this parameter in a given range does not in fact impact the selection due to possibly stronger condition on the EUTRAN cell. Increasing the value of this parameter would: Determine a later selection of GERAN cell which is similar to a shrinking of the GERAN cell. The similar observation made above, regarding the condition that ultimately triggers the selection is applicable for this situation as well. KPI Impact: Mobility - high values might create coverage discontinuity during reselection operation due to shrinking GSM cell as seen by the UE.
Optimization of this parameter, in conjunction with threshServingLow should aim at obtaining the cell sizes for both GERAN cell and EUTRAN cell both in active and in idle mode. Once the cells are correctly dimensioned for active mode, the optimization for idle mode parameters can be performed. The optimization of threshXLow should be decoupled from the optimization for threshServingLow. For this, the value of threshServingLow should be the minimum allowed such that the first inequality of the selection criteria is satisfied for the largest surface of the cell. Once this is realized, the selection will always be triggered by the value of threshXLow. The optimization process should contain the following steps (it is supposed that the sizes of cells in active mode are known): Step 1: Set the value of threshXLow to one of the following values {0, 6, 12, 18, and 24}. Step 2: With UE in idle mode, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another threshXLow and repeat Step 2. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.5
TRESELECTIONGERAN
This parameter concerns the cell reselection timer tReselectionRAT for GERAN. Broadcast in SystemInformationBlockType7. It imposes a condition on the reselection. UE will actually reselect the new cell, only if the new cell is better ranked than the serving cell during a time interval tReselectionGERAN. Recommended & Default Value= "2" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a delayed reselection which could be an issue for fast moving UEs. Decreasing the value of this parameter would: Facilitate ping-pong behaviour during reselection process. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 265/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values of this parameter might allow ping-pong behaviour during reselection operation. High values of this parameter might delay the reselection and possible lead to lost connection to the serving cell. Optimization of this parameter should find a trade-off between delayed reselection and ping pong behaviour. Most probably, if the UEs are not moving fast, the delayed reselection would not be an issue. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionGERAN to one of the following values {1, 2, 3, and 4}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionGERAN. Step 4: Post-process the logs and analyze them as reselection position vs. tReselectionGERAN values and ping pong behaviour vs. tReselectionGERAN values and choose the optimized value to obtain smallest interruption time and highest success rate. Step 5: Calculate the cell reselection success rate in each direction.
15.3.1.6
TRESELECTIONGERANSFMEDIUM
This parameter contributes to the configuration of the IE SystemInformationBlockType7 if the UE is in Medium Mobility state. The concerned mobility control related parameter is multiplied with this factor if the UE is in Medium Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended & Default Value= “0dot5” Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionGERANSfMedium to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionGERANSfMedium. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.7
TRESELECTIONGERANSFHIGH
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 266/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
This parameter contributes to the configuration of the IE SystemInformationBlockType7 if the UE is in High Mobility state. The concerned mobility control related parameter is multiplied with this factor if the UE is in High Mobility state as defined in TS 36.304. This parameter avoids ping pong radio phenomena during the RA-Update & idle mobility. Recommended & Default Value= "oDot25" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier.
KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of tReselectionGERANSfHigh to one of the following values {0.25, 0.5, 0.75, and 1}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the reselection - related messages and the position of the UE. Perform this test 10 times in each direction. Make sure that the driving speed is nominal and the same for all the test samples. Step 3: Repeat Step 2 for another value of tReselectionGERANSfHigh. Step 4: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.8
NCELLCHANGEHIGH
This parameter configures the number of cell changes to enter high mobility state Recommended & Default Value= "12" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. A procedure that optimizes nCellChangeHigh would contain the following steps: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 267/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 1: Set the value of nCellChangeHigh to one of the following values {10,11,12,13,and 14}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.9
NCELLCHANGEMEDIUM
This parameter configures the number of cell changes to enter medium mobility state Recommended & Default Value= "4" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter will determine the UE to start reselection earlier. High values of this parameter will determine the UE to start reselection later. A procedure that optimizes nCellChangeMedium would contain the following steps: Step 1: Set the value of nCellChangeMedium to one of the following values {1, 2, 3, 4, 5, and 6}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another nCellChangeMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.10 QHYSTSFHIGH This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-High included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in High Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection.
Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 268/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
KPI Impact: Mobility - low values of this parameter might allow the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later. A procedure that optimizes qHystSfHigh would contain the following steps: Step 1: Set the value of qHystSfHigh to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfHigh and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.1.11 QHYSTSFMEDIUM This parameter contributes to the configuration of the IE SystemInformationBlockType3.This parameter configures the IE sf-Medium included in the IE SpeedStateReselectionPars. Parameter “Speed dependent ScalingFactor for Qhyst” in TS 36.304. The sf-High concerns the additional hysteresis to be applied, in Medium Mobility state, to Qhyst as defined in TS 36.304 state. This parameter is an environment dependent parameter. This parameter configures the hysteresis value of the serving cell used by the UE for ranking criteria in cell reselection. Recommended & Default Value= "dB-6" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to start cell reselection later. Decreasing the value of this parameter would: Determine the UE to start cell reselection earlier. KPI Impact: Mobility - low values of this parameter might allow the UE to start reselection earlier. High values of this parameter might allow the UE to start reselection later. A
procedure that optimizes qHystSfMedium would contain the following steps:
Step 1: Set the value of qHystSfMedium to one of the following values {dB-6, dB-4, dB-2, dB0}. Step 2: With the UE in idle mode, perform a drive back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE. Step 3: Choose another qHystSfMedium and repeat Step 2. Step 4: Post process the logged data. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 269/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
15.3.2 ACTIVE MODE Both blind redirection and measurement based redirection from LTE to GERAN are supported. LTE to GERAN enhanced redirection that provides system information of the GERAN cells under the redirected carrier to the UE when it is released from LTE and redirected to the GERAN carrier. This functionality can be used for CSFB or non-CSFB redirection from LTE to GERAN
15.3.2.1
REDIRECTION TO GERAN
If eNB receives an event A2 measurement report from UE with measurementPurpose = BelowServing-Floor (UE enters bad air condition area). If the target RAT/carrier selected by eMACT framework is a GERAN carrier, a blind redirection from LTE to GERAN is triggered. If eNB receives an event B2 measurement report from UE with measurementPurpose = MobilityInter-RAT-to-GERAN, but the Cell Change Order to GERAN is not supported by the UE or is not activated in eNB, a measurement based redirection from LTE to GERAN is triggered. Event B2 – Serving becomes worse than threshold1 and inter-RAT neighbour becomes better than threshold2. Entering conditions for this event:
& Ms = measurement result of the serving cell [dBm] Hys = reportConfigGERAN::hysteresis [dB] Thresh1 = ReportConfigGERAN::thresholdEutraRsrpB2 [dBm] Mn = measurement result of the inter-RAT neighbour cell [dBm] Ofn = MeasObjectGERAN::offsetFreq, corresponding to the neighbouring cell [dB] Thresh2 = ReportConfigGERAN::thresholdUtraGeran [dBm]
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 270/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.3-6: Inter RAT threshold for event B2
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 271/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.3-7: Call Flow for Redirection to Geran
15.3.2.2
CELL CHANGE ORDER WITH/WITHOUT NACC TO GERAN
Cell Change Order (CCO) is intending to move a RRC connected UE with packet data session from LTE to GERAN. The trigger is UE measurement-based. eNB will trigger the CCO when UE is leaving LTE coverage and moving into the GERAN coverage, and the UE measurement report indicates the LTE radio condition becomes worse than the value of the configurable parameter thresholdEutraRsrpB2/thresholdEutraRsrqB2 and the GERAN radio condition becomes better than the value of the configurable parameter thresholdGeran. It is include a procedure for eNB to retrieve the system information broadcasted in each of the neighbor GERAN cell through MME. If the information is retrieved successfully, it will be stored in eNB, and will be included in the Cell Change Order when a UE is released from the LTE and moved Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 272/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
to GERAN. The mobility procedure is called Cell Change Order with NACC (Network Assisted Cell Change) if system information of the GERAN target cell is included. Otherwise, the mobility procedure is called Cell Change Order without NACC. For CCO without NACC, UE has to take time to obtain the system information used for call setup by listening to the broadcast channel of the target GERAN cell after UE reselects it. CCO with NACC will speed up the call setup process in target GERAN since system information is already included in the Cell Change Order when UE is released from LTE. CCO is generally used when the optimized inter-RAT PS-handover procedure is not supported by the UE, or by the network. This is because UE has to re-establish connection with GERAN after CCO. However, it is a better mobility procedure than the RRC release/redirection procedure because more GERAN target information can be provided to the UE with CCO and so it will have shorter service interruption time for inter-RAT mobility from LTE to GERAN.
Figure 15.3-8: Serving Radio Condition and UE Measurement Configurations UE is configured to perform the following measurements for the life of the call until UE moves out of the serving cell: a. Event A3 intra-LTE intra-frequency measurement with measurementPurpose = Mobility-IntraFreq). If eNB receives event A3 measurement report from UE, an intra-LTE intra-frequency handover will be triggered to the best cell reported by UE. Intra-LTE intra-Frequency handover trigger has igher priority than any intra-LTE interfrequency or inter-RAT mobility triggers if received simultaneously. b. Event A2 measurements with measurementPurpose = Below-Serving-Floor. If eNB receives event A2 measurement report from UE with measurementPurpose = Below-Serving-Floor, UE enters bad Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 273/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
radio condition (red) as shown in step 3 of Above figure, a blind redirection will be triggered immediately to the target RAT/carrier selected by eMCTA framework. When UE enters good radio condition (green) area, in addition to the life time measurements in step a. and b., eNB will configure and de-configure UE for the following measurement: c. De-configure event A1 measurement with measurementPurpose = Leaving- Coverage-Alarm if UE is currently configured to perform event A1 measurement. d. De-configure all event A5/A3 measurement with measurementPurpose = Mobility-Inter-Freq-toEUTRA if configured. e. De-configure all event B2 measurement with measurementPurpose = Mobility-Inter-RAT-to-UTRA and/or Mobility-Inter-RAT-to-GERAN and/or Mobility-Inter-RAT-to-HRPD if configured f. Configure event A2 measurements with measurementPurpose = Entering-Coverage-Alarm. If eNB receives event A2 measurement report with measurementPurpose = Entering-Coverage-Alarm, UE is moving into alarm radio condition (yellow) area as shown in step 1. When UE enters alarm radio condition (yellow) area, in addition to the life time measurements in step a. and b., eNB will configure and de-configure UE for the following measurements: g. De-configure event A2 measurements with measurementPurpose = Entering-Coverage-Alarm. h. Configure event A1 measurement with measurementPurpose = Leaving-Coverage-Alarm. i. Configure event A5/A3 measurements with measurementPurpose = Mobility-Inter-Freq-to-EUTRA for all intra-LTE inter-Frequency candidates selected by eMCTA framework. j. Configure event B2 measurements with measurementPurpose = Mobility-Inter-RAT-to-UTRA and/or Mobility-Inter-RAT-to-GERAN and/or Mobility-Inter-RAT-to-HRPD for all inter-RAT candidates selected by eMCTA framework. If eNB receives event A5/A3 measurement report for intra-LTE inter-frequency handover or event B2 measurement report for inter-RAT mobility, intra-LTE inter-frequency handover or inter-RAT mobility (PS handover/redirection to UTRA or CCO/redirection to GERAN, or redirection to HRPD) will be triggered. If eNB receives event A1 measurement report with measurementPurpose = Leaving-Coverage-Alarm, UE is moving back into good radio condition (green) area as shown in step 2. In above UE measurement configuration procedure step j, for the GERAN candidates that are selected by eMCTA framework to be measured, eNB will configure UE to perform event B2 measurements with measurement purpose = ‘Mobility-Inter-RAT-to-GERAN’. UE will be triggered to send event B2 measurement report if radio condition of the serving becomes worse than the configurable parameter thresholdEutraRsrpB2/thresholdEutraRsrqB2 and radio condition of the GERAN target becomes better than the configurable parameter thresholdGeran. When the eNB receives a UE event B2 or event B1 (for CS fallback) measurement report for the measurement purpose of ‘Mobility-Inter-RAT-to-GERAN’, if UE supports CCO and ActivationService::isGeranCcoAllowed is set to ‘True’, LTE to GERAN cell change order will be triggered. Otherwise, LTE to GERAN measurement based redirection will be triggered. In above UE measurement configuration procedure step b, if a GERAN carrier is selected as the blind redirection target, LTE to GERAN blind redirection procedured will be trigered
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 274/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Figure 15.3-9: Call Flow for Cell Change Order with /Without NACC
15.3.2.3
THRESHOLDEUTRARSRPB2
This parameter sets the RSRP threshold for the serving cell of the selection criteria in case of CCO with NACC towards GERAN.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 275/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Recommended & Default Value= "-100" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of GERAN cell, i.e. a shrinking of the EUTRAN cell in active mode. Decreasing the value of this parameter would: Determine a later selection of GERAN cell, which is similar with a shrinking of the GERAN cell. KPI Impact: Mobility - low values of this parameter might create coverage discontinuity during selection operation due to shrinking GSM cell as seen by the UE. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of thresholdEutraRsrpB2 to one of the following values {-104,-102,-100,-98,-96}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Step 3: Repeat Step 2 for another value of thresholdEutraRsrpB2. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.4
THRESHOLDGERAN
This parameter sets the RSRP threshold for the target cell of the selection criteria in case of CCO with NACC towards GERAN. Recommended & Default Value= "-110" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine a later selection of GERAN cell, which is similar with a shrinking of the GERAN cell in active mode. Decreasing the value of this parameter would: Determine an earlier selection of GERAN cell, which is similar with a shrinking of the EUTRAN cell. KPI Impact: Mobility - low values of this parameter might create coverage discontinuity during selection operation due to shrinking GSM cell as seen by the UE. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of thresholdGeran to one of the following values {-114,-112,-110,-108, -106,-104}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 276/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 3: Repeat Step 2 for another value of thresholdGeran. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.5
OFFSETFREQGERAN
This parameter is used to indicate a frequency specific offset to be applied when evaluating triggering conditions for measurement reporting. Recommended & Default Value= "0" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine an earlier selection of GERAN cell, which is similar with a shrinking of the EUTRAN cell in active mode. Decreasing the value of this parameter would: Determine a later selection of GERAN cell, which is similar with a shrinking of the GERAN cell. KPI Impact: Mobility - low values of this parameter might allow the UE to determine the strongest cell later. High values of this parameter might allow the UE to determine the strongest cell earlier. For optimization, a procedure containing the following steps can be used: Step 1: Set the value of offsetFreqGERAN to one of the following values {-3,-2,-1, 0, 1, 2, 3}. Step 2: While performing a download with the UE, perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the control messages exchanged between eNodeB and UE along with the GPS coordinates of the UE Step 3: Repeat Step 2 for another value of offsetFreqGERAN. Step 4: Post process the logged data and determine the positions at which the UE selected the GERAN cell. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.6
FILTERCOEFFICIENTOFQUANTITYCONFIGGERAN
This parameter is used to configure the IE filterCorefficient of QuantityConfigGERAN. The parameter is optional and is required only when inter-RAT mobility to GERAN is supported. If this parameter is not configured (absent) then the default RRC value defined in 36.331 is used by the eNB and signalled to the UE. The RSRP values reported by the UE are obtained by filtering several measurements performed by the UE. If this filter can allow quick variation to be reported or it can rely more on the last reported value and less on the measured value such that there is less variation in the sequence of the reported value. The higher the value of filterCoefficientOfQuantityGERAN the smoother the reported measurement will be and consequently the less likely ping-ponging occurs between sectors during handover. Recommended & Default Value= "fc2" Expected behaviour when changing this parameter: Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 277/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Increasing the value of this parameter would: Decrease the variation in the reported RSRP value. Decrease ping-pong between the cells in case of handover conditions. Delay the speed at which the reported RSRP adapts to the RSRP variation. This might eventually slightly delay the HO, if the value of the parameter is too high. Improve the system behaviour regarding the throughput during HO. Decreasing the value of this parameter would: Increase the variation in the reported RSRP value due to noise. Increase the ping-pong between the cells in case of handover conditions due to variations in reported RSRP. Decrease the HO quality relative to throughput. Increase the speed at which the reported RSRP adapts to the RSRP variation. This might speed up the HO which could manifest as ping-pong.
KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Optimization of this parameter should be performed in conjunction with optimization of hysteresis and timeToTrigger parameters. Finding the optimum pair of (filterCoefficientOfQuantityGERAN, hysteresis, timeToTrigger) should consider the following steps: Step 1: Set the values of filterCoefficientOfQuantityGERAN and to hysteresis and to timeToTrigger to one of the following {(fc2, 3,100), (fc3, 4, 80), (fc4, 5,200), (fc1, 2, 40)}, in both current cell and neighbour cell. Step 2: Perform a drive test while performing a download and log the throughput values and the position of the UE. Drive in and out of the current cell to the neighbour cell. Step 3: Repeat Step 2 for another pair of values of the three tested parameters. Step 4: Represent throughput vs. position (distance) (Service continuity), #HO-attempts, Success Rate/Failure Rate, #of Ping-pongs, HO interruption time for all pairs of tested values. Step 5: Choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.7
HYSTERESIS
This IE is a parameter used within the entry and leave condition of an event triggered reporting condition. This is used to provision IE Hysteresis in IE ReportConfigInterRAT, in IE MeasConfig . This parameter defines the hysteresis used by the UE to trigger an inter-RAT event-triggered measurement report. It is used in several processes: Event B2 (Serving becomes worse than threshold1 and inter RAT neighbour becomes better than threshold2); Event B1 (Inter RAT neighbour becomes better than threshold); Event A1 (Serving becomes better than threshold); Event A2 (Serving becomes worse than threshold); Event A3 (Neighbour becomes offset better than serving); Event A4 (Neighbour becomes better than threshold); Event A5 (Serving becomes worse than threshold1 and neighbour becomes better than threshold2). Recommended & Default Value= "3" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO due to the more important difference that must exist between the serving cell and neighbour cell. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 278/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Drop the call if the value is too large i.e. connection to the serving cell is lost before having reached the neighbour cell level that satisfies the HO condition. Decreasing the value of this parameter would: Create a ping – pong behaviour because the measurement quick variations (noise-like) might trigger HO decisions. KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. Throughput - low values of this parameter can generate a ping pong behaviour which can result in interruption times and low throughput during HO operation.
Optimization of this parameter should be performed in conjunction with optimization of filterCoefficientOfQuantityGERAN and timeToTrigger parameters, as presented in the previous paragraph.
15.3.2.8
TIMETOTRIGGER
This parameter sets the time duration time during which the conditions to trigger an event report have to be satisfied before sending a RRC measurement report in event triggered mode. Recommended & Default Value= "ms100" Expected behaviour when changing this parameter Increasing the value of this parameter would: Delay the HO decision. Determine a call drop due to significant serving cell signal degradation before timeToTrigger expires. Decreasing the value of this parameter would: Generate ping-pong HO behaviour due to the fact that quick variations of the measured signal (noise-like variations) might satisfy the HO relation for the short while represented by timeToTrigger but not much longer. KPI Impact: Mobility – low values of this parameter might allow ping-pong behaviour during HO operation. High values of this parameter might delay the HO and possible lead to lost connection to the serving cell. This parameter should be carefully optimized, best in conjunction with filterCoefficientOfQuantityGERAN and hysteresis as presented in paragraph 15.3.2.6 and 15.3.2.7. Indeed, the optimized value can be impacted by the load of the surrounding cells.
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 279/290
LTE Optimization Handbook TLA6.0
15.3.2.9
MGR/TIPS/NEA
REPORTINTERVAL
This parameter configures the IE reportInterval included in the IE ReportConfigInterRAT in the MeasConfig IE. The ReportInterval indicates the interval between periodical reports. The ReportInterval is applicable if the UE performs periodical reporting (i.e. when reportAmount exceeds 1), for triggerType ‘event’ as well as for triggerType ‘periodical’. Recommended & Default Value= "ms240" Expected behaviour when changing this parameter Increasing or decreasing the value of this parameter would: Will not decrease or increase the Handover Success rate; but no point to have high values.
KPI Impact: Mobility – low values of this parameter will decrease the HO success rate. High values of this parameter will increase the Ho success rate. A procedure that optimizes reportInterval would contain the following steps: Step 1: Set the value of reportInterval to one of the following values {120, 240, 480, 640, 1024, and 2048}. Step 2: perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportInterval. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.10 MAXREPORTCELLS This parameter defines the maximum number of cells to be reported in a measurement report. Recommended & Default Value= "1" Expected behaviour when changing this parameter Increasing the value of this parameter would: Determine the UE to report as many new neighbour cells as possible in a short time. Decreasing the value of this parameter would: Determine the UE to report fewer neighbour cells. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells. A procedure that optimizes maxReportCells would contain the following steps: Step 1: Set the value of maxReportCells to one of the following values {1, 2, 3, 4, 5, 6, 7, 8}. Step 2: Perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of maxReportCells. Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 280/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
15.3.2.11 REPORTAMOUNT This parameter configures the number of periodical reports the UE has to transmit after the event was triggered. Recommended & Default Value= "r8" Expected behaviour when changing this parameter Increasing the value of this parameter would: Increase the Handover Success Rate for multiple repetitions in bad RF conditions. Decreasing the value of this parameter would: Decrease the Handover Success Rate for multiple repetitions in bad RF conditions. KPI Impact: Mobility – low values of this parameter allow the UE to report fewer neighbour cells. High values of this parameter allow the UE to report more neighbour cells. A procedure that optimizes reportAmount would contain the following steps: Step 1: Set the value of reportAmount to one of the following values {r1, r2, r4, r8, r16, r32, r64}. Step 2: Perform a drive test back and forth between the EUTRAN cell and GERAN cell on various routes and log the HO - related messages and the position of the UE. Step 3: Repeat Step 2 for another value of reportAmount. Step 4: Post process the measurement and choose the optimized value to obtain smallest interruption time and highest success rate.
16 ABBREVIATIONS AND DEFINITIONS 16.1 ABBREVIATIONS Acronym 3G
Description 3rd Generation Mobile Telecommunications
3GPP
3rd Generation Partnership Project
3GPP2
EV-DO standards
AMBR
Aggregate Maximum Bit Rate
AMR
Adaptive Multi Rate codec
ANR
Automatic Neighbour Relation
ASN.1
Abstract Syntax Notation 1
ASN1
Abstract Syntax Notation One
BBU
Base Band Unit, D-BBU, H-BBU for HSDPA, E-BBU for HSUPA
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 281/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
BCCH
Broadcast Control Channel
BCH
Broadcast Channel
BLER
Block Error Rate
BRC
Baseband Resources Controller
CAC
Connection Admission Control
CallP
Call Processing
CCCH
Common Control Channel
CCM
Channel Control Module
CDMA
Code Division Multiple Access
CGI
Cell Global Identification = MCC + MNC + LAC + CI
CH
Channel
CI
Cell Identity
CK
Cipher Key
CLR
Cell Loss Ratio
CM
Configuration Management
CMIP
Client Mobile IP
CN
Core Network
cNode
Control Node
CoS
Class of Service
CQI CR
Channel Quality Indicator (UE transmits a CQI report at regular intervals indicating the current DL radio conditions) Change Request (a problem report within the Clarify system)
CRNC
Controlling Radio Network Controller
C-RNTI
Cell RNTI (16 bits)
CS
Circuit Switch
CT
Call Trace
CTCH
Common Traffic Channel
CTS
5420 CTS – Converged Telephony Server (formerly Feature Server 5000)
CUM
Cumulative counter
D2U DCCH
Base Band Unit (BBU) – d2U and d1U; A signal Distributed 2U (d-2U) digital unit, this indoor unit contains the channel element cards and the control module. Dedicated Control Channel
DCH
Dedicated Channel
DCT
Defect & Change Tracking tool
DFT
Discrete Fourier Transform
DHCP
Dynamic Host Configuration Protocol
DL
Downlink (equivalent to EV-DO Forward Channel) LTE supports peak data rate of 300Mbps DL
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 282/290
LTE Optimization Handbook TLA6.0 DL-SCH DLU
MGR/TIPS/NEA
Downlink Shared Channel BTS MIB
DRB
Data Radio Bearer
DRNC
Drift Radio Network Controller
DSCH
Downlink Shared Channel
DSMIP
Dual Stack Mobile IP
DSMIPv6
Dual Stack MIPv6
DTCH
Dedicated Traffic Channel
E1
Standard European PCM link nickname
E911
Enhanced 911
EBI
EPS Bearer Id
ECM
EPS Connection Management
EIR
Equipment Identity Register
EMM
EPS Mobility Management (part of NAS)
eNB
Evolved NodeB (or eNodeB) (combines functions of UMTS NodeB and RNC)
EPC
Evolved Packet Core
ePLMN
Equivalent Public Land Mobile Network
EPS
Evolved Packet System
ESM
EPS Session Management (part of NAS)
EUTRAN
Evolved UMTS Terrestrial RAN
E-UTRAN
Evolved Universal Terrestrial Radio Access Network
FACH
Forward access Channel
FACT
First Acceptance Criteria Test
FCAPS
Fault, Configuration, Accounting, Performance, and Security (OAM term)
FDD FFS
Frequency Division Duplex (UE operates on one frequency for UL and another frequency for DL) For Further Study
FFT
Fast Fourier Transform (one split into many)
FM
Fault Management
FOA
First Office Application
FTP
File Transfer Protocol
GBR
Guaranteed Bit Rate
GGSN
Gateway GPRS Support Node
GPRS
General Packet Radio Service
GSM
Global System for Mobile communications
GTP
GPRS Tunnelling Protocol
GTP-PDU
GTP-C PDU or GTP-U PDU
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 283/290
LTE Optimization Handbook TLA6.0 GUMMEI
MGR/TIPS/NEA
HARQ
Globally Unique MME Identifier = MCC + MNC + MMEI = MCC + MNC + MME Group Id + MMEC Globally Unique Temporary Identifier = GUMMEI + M-TMSI = MCC + MNC + MME Group Id + MMEC + M-TMSI Hybrid Automatic Repeat Request
HHO
Hard Hand Over
HLD
High Level Design
HLR
Home Location Register
HO
Handover
HoA
Home IP Address
H-PCEF
A PCEF in the HPLMN
hPLMN
home Public Land Mobile Network
HRPD
High Range Packet Data
HSGW
HRPD Serving Gateway
HW or H/W
Hardware
IBTS
Internet BTS
ICI
Inter-Carrier Interference
ICS
IMS Centralized Services
IFFT
Inverse Fast Fourier Transform (many combined into one)
IMA
Inverse Multiplexing for ATM
IMEI
International Mobile Equipment Identity
IMEISV
International Mobile Equipment Identity with Software Version number
IMS
IP Multimedia Subsystem
IMSI
International Mobile Station Identifier = MCC + MNC + MSIN
IN or iNode
Interface Node
Inter-RAT HO Inter-System HO IOT
Inter Radio Access Technology handover (UMTS-GSM)
IP
Internet Protocol
IP-CAN
IP Connectivity Access Network
IPv4
IPv4 IP Address: e.g., 135.2.80.116
IPv6 ISI
IPv6 IP Address: e.g., 002:00D3:0000:0000:02AA:0000:FE28:9C5A OR 2:D3:0:0:2AA:0:FE28:9C5A (delete leading zero’s) OR 2:D3::2AA:0:FE28:9C5A (onetime collapse of one/multiple zero’s) Inter-Symbol Interference
ISS
Integration SubSystem team
ITP
Integration Test Plan
Iu
CN-UTRAN interface
GUTI
Inter System Handover, 3G to 2G or 2G to 3G Inter Operability Test
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 284/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
Iub
Interface between RNC and NodeB
Iucs
Iu Circuit Switch
Iups
Iu Packet Switch
Iur
Interface between two RNCs
KPI
Key Performance Indicator
L1
Layer 1
L2
Layer 2
L3
Layer 3
LAC
Location Area Code (2 octets)
LAI
Location Area Identifier = MCC + MNC + LAC
LBI
Linked EPS Bearer Id
LI
Lawful Intercept
LLDM
LGE UE Diagnostic Monitor
LMA
Local Mobility Anchor
LMD
Local Mobility Domain
LTE
Long Term Evolution
MAC
Media Access Control
MAG
Mobile Access Gateway
MAP
Mobility Anchor Point (MIP)
MBR
Maximum Bit Rate
MCC
Mobile Country Code (3-digits) (Ref: ITU-T Rec E.212,AnnexA)
MCCH
Multicast Control Channel
MEI
Mobile Equipment Identity
MIB
Master Information Block
MIM
Management Information Model
MIMO
Multiple Input Multiple Output antenna technique
MIP
Mobile IP
MIPv4
Mobile IPv4
MIPv6
Mobile IPv6
MME
MMEI
Mobility Management Entity (Mobility management functions, paging authentication, S-GW selection, PDN-GW selection) Mobility Management Entity Group Identifier (16 bits) (identifies the MME Pool to which an MME belongs) Mobility Management Entity Code (8 bits) (identifies a MME within the scope of a MME GroupId in a PLMN) or (uniquely identify a MME within a MME pool area) Mobility Management Entity Identifier = MMEGroupId + MMEC
MNC
Mobile Network Code (2 or 3-digits) (Ref: Figure 10.5.154 of 3GPP TS 24.008)
MO
Managed Object
MME GroupId MMEC
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 285/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
MSC
Mobile Switching Centre
MSIN
Mobile Subscriber Identification Number (used in IMSI)
MSISDN
Mobile Subscriber international PSTN/IDSN number = CC + NDC + SN
MTA
Mobile Trace Analyzer tool
MTCH
Multicast Traffic Channel
M-TMSI N/A
MME Temporary Mobile Subscriber Identifier (32 bits) (Allocated by MME) (unique identifier for UE within MME) Not Applicable
NAS
Non-Access Stratum
NBAP
Node B Application Part
NDC
National Destination Code
NE
Network Element
NMSI
National Mobile Subscriber Identity = MNC + MSIN
NNI
Network-Node Interface
Node B
Base Transceiver Station
NSAP
Network Service Access Point
OAM
Operations and Maintenance
OC3
Optical Carrier-3 (155.52 Mbit/s)
OCAN
Offline Configuration for Access Network
OCS
Online Charging System
OFCS
Offline Charging System
OFDM
Orthogonal Frequency Division Multiplexing
OFDMA
Orthogonal Frequency Division Multiple Access – DL air interface
OLC
One Logical Cell (may also named “supper cell”)
OMC-B
Operation and Maintenance Centre for NodeB
OMC-P
Operations Management Console - Provisioning
OMC-R
Operation and Maintenance Centre for RNC
OMU
Operation and Maintenance Unit
OTSR
Omni Transmit Sector Receive
PA
Power Amplifier
PAA
PDN Address Allocation
PBCH
Physical Broadcast Channel
PCC
Policy and Charging Control
PCCH
Paging Control Channel
PCEF
Policy Charging Enforcement Function
PCH
Paging Channel
PCO
Protocol Configuration Option (for PMIP Binding Acknowledgement)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 286/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
PCR
Peak Cell Rate
PCRF
Policy Charging Rule Function
PDCCH
Physical Downlink Control Channel
PDCP
Packet Data Convergence Protocol
PDM
Packet Data Monitoring tool
PDN
Packet Data Network
PDN-GW PDSCH
Packet Data Network Gateway (or P-GW) (IP address allocation, Policy enforcement, Packet filtering) Physical Downlink Shared Channel
PDTI
Plan de Tests d’Intégration (Integration test plan)
PDU
Protocol Data Unit
PLMN
Public Land Mobile Network = MCC + MNC
PM
Performance Management
PMIP
Proxy Mobile IP
PMIPv6
Proxy Mobile IPv6
PMK
Pairwise Master Key
PNNI
Private Network-Network Interface
POR
Plan Of Record
PPPMT
PPP Monitoring Tool (now Packet Data Monitoring tool)
PRACH
Physical Random Access Channel
PRB
Physical Resource Blocks
PS
Packet Switched
P-SCH
Primary synchronization channel
PTI
Protocol / Procedure Transaction Id
PUCCH
Physical Uplink Control Channel
PUSCH
Physical Uplink Shared Channel
QC
Quality Center
QCI
QoS Class Identifier
QoS
Quality of Service
QRM
Quality and Reliability Measurements (PM XML file containing QRM counters)
RAC
Routing Area Code (2 octets)
RACH
Random Access Channel
RAI
Routing Area Identification = MCC + MNC + LAC + RAC
RAN
Radio Access Network
RANAP
RAN Application Part
RAT
Radio Access Technology
RB
Radio Bearer
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 287/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
RF
Radio Frequency
RLC
Radio Link Control
RNC
Radio Network Controller
RNTI
Radio Network Temporary Identifier
RO
Resource Object
RRC
Radio Resource Control (3GPP TS 36.331)
RRH
Remote Radio Head
RRM
Radio Resources Management
RS
Reference Symbol
RSRP
Reference Signal Received Power
RSRQ
Reference Signal Received Quality
RSSI
Received Signal Strength Indicator
RSVP
Resource Reservation Protocol
S1AP
S1 Application Protocol
S1-U
Interface between SGW and eNodeB
SAAL-NNI
Signaling ATM Adaptation Layer - Network Node Interface
SAC
Service Area Code (2 octets)
SAE
System Architecture Evolution
SAI
Service Area Identification = MCC + MNC + LAC + SAC
SAP
Service Access Point
SAR
Segmentation And Reassembly
SB
Scheduling Block
SCCP
Signaling Connection Control Part
SC-FDMA SCTP
Single Carrier Frequency Division Multiple Access – UL air interface (OR DFT-Spread OFDMA) (results in very low Peak-to-Average Power Ratio (PAPR)) Stream Control Transmission Protocol (used on S1-AP, X2-AP interfaces)
SDF
Services Data Flow
SDH
Synchronous Digital Hierarchy
SDM
Services Data Manager
SDMA
Space Division Multiple Access (antenna)
SectorID
Sector Address Identifier
SFN
System Frame Number
SGSN
Serving GPRS Support Node
SGW
Signalling Gateway
S-GW SHO
Serving Gateway (User plane anchor point for inter-eNodeB handovers and inter3GPP handovers) Soft Hand Over
SIB
System Information Block
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 288/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
SINR
Signal to Interference-plus-Noise Ratio
SM
Security Manager
SN
Subscriber Number
SNR
Signal to Noise Ratio
SOAP
SON
Simple Object Access Protocol (a lightweight protocol that is commonly used to send XML messages over the Internet) Scalable Orthogonal Frequency Division Multiple Access (keeps subcarrier spacing constant; better for handovers) Self Organizing Network
SONET
Synchronous Optical Network
SPR
Subscription Profile Repository
SR
Scheduling Request
SRB
Signaling Radio Bearer
SRLR
Synchronized Radio Link Reconfiguration
SRNC
Serving Radio Network Controller
SRNS
Serving Radio Network System
SRS
Sounding Reference Signal
SSCF
Service Specific Co-ordination Function
S-SCH
Secondary Synchronization Channel
SSCOP
Service Specific Connection Oriented Protocol
SSCS
Service Specific Convergence Sub layer
SSSAR
Service Specific Segmentation and Re-assembly sub layer
STI
Spécification des Tests d’Intégration (Intégration test spécification)
STM1
Synchronous Transport Module-1 (155.52 Mbit/s)
S-TMSI
Serving Temporary Mobile Subscriber Identifier = MMEC (8 bits) + M-TMSI (32 bits)
SOFDMA
STSR1
Sectorized Tx Sectorized Rx – 1 frequency
STSR2
Sectorized Tx Sectorized Rx – 2 frequency
SU
Scheduling Unit
SW or S/W
Software
TA
Tracking Area
TAC
Tracking Area Code
TAI
Tracking Area Identity = MCC + MNC + TAC
TAU
Tracking Area Update
TBM
Transport Bearer Management
TC
Test Case
TCP/IP
Transmission Control Protocol/Internet Protocol
TDD
Time Division Duplex (UE operates on one frequency with different time slots for UL and DL)
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 289/290
LTE Optimization Handbook TLA6.0
MGR/TIPS/NEA
TEID
Tunnel Endpoint Identifier (GTP)
TEID-C
Tunnel Endpoint Identifier, Control Plane
TEID-U
Tunnel Endpoint Identifier, User Plane
TI
Transaction Identifier
TIL
Terminal d’Installation Locale (Local I&C terminal)
TM
Transparent Mode
TMSI
Temporary Mobile Station Identifier
TMU
Traffic Management Unit
TRB
Traffic Radio Bearer
TS
Technical Specification
UBM
Upper Bearer Management
UDP
User Datagram Protocol
UE
User Equipment (similar to Access Terminal in EV-DO)
UEA
UMTS Encryption Algorithm
UICC
Universal Integrated Circuit Card
UIIV
UMTS Interoperability Integration and Verification
UL ULI
Uplink (equivalent to EV-DO Reverse Channel) LTE supports peak data rate of 80 Mbps UL User Location Info
UL-SCH
Uplink Shared Channel
UMTS
Universal Mobile Telecommunication System
UTRAN
Universal Terrestrial Radio Access Network
VAL
Value counter
VLR
Visitor Location Register
VoA
Visited IP Address
VoIP
Voice over Internet Protocol
VSNCP
Vendor Specific Network Control Protocol
WCDMA
Wideband CDMA
WPS or WiPS
9452 Wireless Provisioning System (also sometimes referred to as WIPS)
END OF DOCUMENT
Alcatel-Lucent - Confidential - Solely for authorized persons having a need to know Proprietary - Use pursuant to Company instruction
LTE/IRC/APP/032749
V06.03 / EN
Approved Standard
28/Oct/2013
Page 290/290