LTE Trial Test Result (Etisalat) -V1.6

LTE/SAE Trial Results Etisalat UAE Nokia Siemens Networks LTE Trial Team July 2010, Dubai UAE

1

© Nokia Siemens Networks

Contents • Network Configuration • Coverage Plot based on drive test • Single User Throughput • Sector Throughput • Latency • Mobility: cell reselection and Handover • Link Budget • Annex: eNB automatic neighbors list creation

LTE Trial for Etisalat – Test results 2


Customer Confidential

Network Configuration




Coverage Plot Based on Drive Test




Condition: • One UE with TCP Downlink application • Only site 4106 is unlocked, the other sites are locked • TX power was set as 60 watt • Logging with XCAL Drive Test tool




Serving Cells Plot




RSRP Coverage Plot (dBm)




SINR Coverage Plot (dB)




Downlink Throughput (Mbps)




Single User Throughput




Single User Throughput – Test Scenario The objective of the test is to verify the capability of the LTE air interface to provide DL and UL throughput for a single user in different radio conditions with different application protocols

• Application protocols – TCP (DL/UL) – UDP (DL/UL) • DL antenna configuration – 2 x 2 DL Adaptive OL MIMO • Radio conditions – stationary • UE is LTE category 3 Samsung device (Throughput is limited 100/50 Mbps (DL/UL)) • UE device providing 1 second measurement period to Xcal LTE tool




Single User Throughput ‐ TCP Downlink • Average RSRP : ‐70.71 dBm, Average SINR : 27.67 dB • DL average throughput : 90.16 Mbps • DL peak throughput : 101.1 Mbps *

* UE device providing 1 second measurement period to Xcal LTE tool LTE Trial for Etisalat – Test results 12



Single User Throughput ‐ TCP Uplink • Average RSRP : ‐69.74 dBm, Average SINR : 26.76dB (Both measured in DL) • UL average throughput : 36.50 Mbps • UL peak throughput : 40.48 Mbps




Single User Throughput ‐ UDP Downlink • Average RSRP : ‐65.35 dBm, Average SINR : 26.24 dB • DL average throughput : 93.90 Mbps • DL peak throughput : 102.11 Mbps * Phy DL TP(Mbps) 120 100 80 Phy DL TP(Mbps)

60 40 20

* UE device providing 1 second measurement period to Xcal LTE tool LTE Trial for Etisalat – Test results 14



17:16:58.366

17:16:53.052

17:16:46.365

17:16:40.364

17:16:35.364

17:16:30.362

17:16:25.362

17:16:20.361

17:16:14.360

17:16:09.052

17:16:03.362

17:15:58.360

17:15:53.359

17:15:48.358

17:15:43.052

17:15:37.357

17:15:32.356

17:15:27.355

17:15:21.358

17:15:16.052

17:15:10.354

0

Single User Throughput – UDP Uplink • Average RSRP : ‐69.39 dBm, Average SINR : 26.92 dB (Both Measured in DL) • UL average throughput : 35.94 Mbps • UL peak throughput : 38.76 Mbps




Single User Throughput ‐ TCP Downlink Summary

Measured Radio condition Reference LTE Trial for Etisalat – Test results 16



Single User Throughput ‐ TCP Uplink Summary

Measured Radio condition * RSRP & SINR measured in DL LTE Trial for Etisalat – Test results 17


Reference Customer Confidential

Single User Throughput ‐ UDP Downlink Summary

Measured Radio condition Reference LTE Trial for Etisalat – Test results 18



Single User Throughput ‐ UDP Uplink Summary

Measured Radio condition * RSRP & SINR measured in DL

Reference

Poor condition was not achieved during the test due to sudden increase in the SINR during testing session LTE Trial for Etisalat – Test results 19



Sector Throughput




Objectives & Conditions •Sector throughput – Scenario A.1 – Four UEs placed stationary in excellent radio conditions – TCP DL/UL traffic (FTP) – Cell capacity is the aggregated physical layer throughput from each UE •Sector Throughput – Scenario A.2 – Four UEs placed stationary in good radio conditions – TCP DL/UL (FTP) – Cell capacity is the aggregated physical layer throughput from each UE • Sector throughput – Scenario B – Two UEs placed stationary in good radio conditions, one in average and one in poor radio conditions – TCP DL/UL traffic (FTP) – Cell capacity is the aggregated physical layer throughput from each UE •Sector throughput – Scenario C – Three UEs placed stationary in good radio conditions and one UE mobile (from good to average and poor, then back to good) – TCP DL/UL traffic (FTP) – Cell capacity is the aggregated physical layer throughput from each UE




Sector Throughput Downlink Direction




Scenario A.1 ‐ TCP DL Results • • • •

Two Stationary UEs in Excellent Radio Condition and another two stationary UEs in Good radio condition UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 118.1 Mbps Peak Sector Throughput 142.76 Mbps Reference



Measured Radio condition


Scenario A.2 ‐ TCP DL Results • • • •

Four Stationary UEs in Good Radio Conditions UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 78.25 Mbps Peak Sector Throughput 95.66 Mbps

Reference





Scenario B ‐ TCP DL Results • • • •

UE1 in Average RF, UE2 in Poor RF, UE3 & UE4 in Good RF UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 61.73 Mbps Peak Sector Throughput 99 Mbps

Reference





Scenario C - TCP DL Results 1/2 • • • •

UE1, UE3, UE4 in Good RF. UE2 is mobile moving from Good RF to Average and to Poor then return back to Good UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 105.05 Mbps Peak Sector Throughput 118.1 Mbps Reference





Scenario C - TCP DL Results 2/2 •

GIS Plot of Moving UE2 can be seen as below




Sector Throughput Uplink Direction




Scenario A.1 ‐ TCP UL Results • • • •

Two Stationary UEs in Excellent Radio Condition and another two stationary UEs in Good radio condition UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 34.87 Mbps Peak Sector Throughput 43.98 Mbps Reference





Scenario A.2 ‐ TCP UL Results • • • •

Four Stationary UEs in Good Radio Conditions UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 24.06 Mbps Peak Sector Throughput 32 Mbps

Reference





Scenario B ‐ TCP UL Results • • • •

UE1 in Average RF, UE2 in Poor RF, UE3 & UE4 in Good RF UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput Average Sector Throughput 21.11 Mbps Peak Sector Throughput 32.2 Mbps

Reference





Scenario C - TCP UL Results 1/2 • UE1, UE3 and UE4 are in excellent or good RF, UE2 is mobile from good to average and poor RF • UEs Synchronized with NTP server before the test and results aggregated with time stamps to get sector throughput • Average Throughput 18.6 Mbps • Peak Throughput 32.5 Mbps Reference





Scenario C - TCP UL Results 2/2 •

GIS Plot of Moving UE2 can be seen as below




Latency




Objectives & Conditions • Objective: Measure the control plane signaling processing time and the end-to-end user plane delay, i.e. Round Trip Time (RTT). Tests will be executed in the field at good SINR conditions. • Control plane: – Location in field: stationary UE’s in the target cell at good SINR locations – Signaling procedures: UE Attach, Radio Bearer Setup and UE Detaches from active state • User plane: – Location in field: stationary UE’s in the target cell at good SINR locations – Ping application server with 32Bytes, 1000Bytes, 1500Bytes with 100 pings session




Control Plane Latency – Attach & Bearer Setup

29 ms 5 ms

116 ms

5 ms

Total time for UE to attach (EMM-registered) and setup default bearer (ECM-Connected) takes around 155 ms (including UE processing delay). LTE Trial for Etisalat – Test results 36



Control Plane Latency – UE Detach

46 ms

1 ms

Total time of UE detach from active state takes around 47 ms. LTE Trial for Etisalat – Test results 37



User Plane Latency-32Bytes Ping Average: 11ms Minimum: 10ms Maximum: 33ms












User Plane Latency - Summary 60 50

RTT (ms)

40 27

30

23

20 11

10 0 Average

Minimum

Maximum 32B




1000B

1500B

Mobility: Cell Re-selection




Objectives & Conditions • Objective: Verify that cell reselection happens within the conditions set by the parameters. • Parameters: – Sintrasearch – qrxlevmin – qoffset – qhyst – Treselection




Cell reselection Process

UE will start to measure intra frequency neighbors when Rx level of serving cell is below sum of Sintrasearch + qrxlevmin. Cell reselection will happen when Rn > Rs within Treselection timer period. LTE Trial for Etisalat – Test results 44



Cell Reselection Result

BCCH-DL SCH information list cell 101 as neighbor for cell 100

UE camp on cell 100. When the RSRP of cell 101 is 3dB (qoffset + qhyst) higher than cell 100 then cell reselection is triggered if this condition is fulfilled for 2 seconds (Treselection).




Mobility: Handover




Objectives & Conditions • Objective: Verify functionality of intra-eNodeB and inter eNodeB triggered by the event A3 (better cell HO) and A5 (handover threshold based on RSRP measurements).

• Application: TCP IP DL and UL

• Location: UE performing intra/inter eNodeB handover from serving to target cell/target eNodeB

• Vehicle Speed: The range of speed during the handover tests was between 30 to 60 Kph




Mobility and Handover – Condition The successful handover was determined by verifying the following message sequence: • Measurement Report sent from UE to source cell. – Event A3: Neighbouring Cell becomes better than Serving Cell. – Event A5: Serving becomes worse than threshold1 and neighbor becomes better than threshold2

• RRCConnectionReconfiguration message including MobilityControlInfo sent by a source cell to UE. • RRCConnectionReconfigurationComplete sent from UE received by a target cell The control plane interruption time was determined by calculating: • The time from a UE receiving the RRCConnectionReconfiguration message in the source cell to sending the RRCConnectionReconfigurationComplete message in the target cell.




Mobility and Handover – Intra eNodeB time

35 ms

Handover command 29 ms

Handover complete

Total time from measurement report to handover: 64 ms (in average) LTE Trial for Etisalat – Test results 49



Handover Events Identification

In rrcConnectionReconfiguration message for measurement configuration, eNodeB sent reportConfigId identifier. In the measurementReport message, measId identifier refer to the handover event: - measId 1 = event A3 - measId 2 = event A5 - measId 3 = event A2 LTE Trial for Etisalat – Test results 50



Handover procedure triggered by event A3




Mobility and Handover – Intra eNodeB with A3 HO trigger

Serving cell RSRP = 49, i.e. -140 + 52 = -91 dBm

Target cell RSRP = 52, i.e. -140 -52 = -88 dBm

The Event A3 reporting threshold condition (eNB parameter: a3Offset = 3 dB) is fulfilled and the UE sends handover measurement report to the serving eNodeB. The intra eNodeB handover from Cell ID 102 to Cell ID 100 is executed. LTE Trial for Etisalat – Test results 52



Handover procedure triggered by event A5




Mobility and Handover – Intra eNodeB with A5 HO trigger



In order to perform HO with A5 trigger, A3offset has been changed to avoid HO due to A3 (15 dB). Based on measurement report, it is showing that UE was served by cell 100. And measure cell 101 as target when serving cell below A5 threshold (set by parameter was -80 dBm) and decided to HO since the target is better than -77 dBm (set by parameter) after time to trigger is achieved. LTE Trial for Etisalat – Test results 54



Intra eNodeB Handover – DL Throughput Impact (1 of 2) Number of Handover attempts were 14, 7 in clockwise and 7 in counterclockwise, Handovers with 100% success rate

Handovers were performed from cell ID 102 to 100, 100 to 101 and 101 to 102. During the handover, there is no throughput gap. Fluctuations in throughput are due to SINR during mobile test. LTE Trial for Etisalat – Test results 55



Intra eNodeB Handover– DL Throughput Impact (2 of 2)

Fluctuation of throughput during handover is due to SINR. There is no gap throughput during handover. LTE Trial for Etisalat – Test results 56



Intra eNodeB Handover– UL Throughput Impact (1 of 2) Number of Handover attempts were 16, 8 in clockwise and 8 in counterclockwise, Handovers with 100% success

Handovers were performed from cell ID 101 to 102, 102 to 100 and 100 to 101. During the handover, there is no throughput gap. Fluctuations in throughput are due to SINR. LTE Trial for Etisalat – Test results 57



Intra eNodeB Handover– UL Throughput Impact (2 of 2)

Fluctuation of throughput during handover is due to SINR. There is no gap throughput during handover. LTE Trial for Etisalat – Test results 58



Mobility and Handover – Inter eNodeB

53 ms

Handover command 30 ms

Handover complete

Total time from measurement report to handover: 83 ms (in average) LTE Trial for Etisalat – Test results 59



Mobility and Handover – Inter eNodeB



UE was originally camping to cell 401 belong to site 2549. Due to A3 HO trigger, UE handover to cell 100 belong to site 4106. LTE Trial for Etisalat – Test results 60



Inter eNodeB Handover– DL Throughput Impact (1 of 2) Number of Handover attempts is 20, Handovers with 100% success rate

Handovers were performed between cells ID 100 to 202 and 400. During the handover, there is no throughput gap. Fluctuations in throughput are due to SINR (see next slide). Speed was between 30 and 60 Kph. LTE Trial for Etisalat – Test results 61



Inter eNodeB Handover– DL Throughput Impact (2 of 2)




Inter eNodeB Handover– UL Throughput Impact (1 of 2) Number of Handover attempts is 16, Handovers with 100% success rate

Handovers were performed between cells ID 100 to 202 and 400. During the handover, there is no throughput gap. Fluctuations in throughput are due to SINR (see next slide). Speed was between 30 and 60 Kph. LTE Trial for Etisalat – Test results 63



Inter eNodeB Handover– UL Throughput Impact (2 of 2)




Link Budget




Link Budget – Test Scenario • Objective: To show relation between achieved throughput with SINR and RSRP

• Application: TCP IP downlink and uplink

• Location: One UE using applications (TCP DL, TCP UL) move from good SINR to cell edge (poor SINR) for antenna configuration (adaptive OL MIMO DL), without interference (all neighboring cells with 0% load). Cell edge will be searched and the UE will be moving along the predicted cell edge to have statistically reliable results




Link budget – RSRP versus DL throughput




Link budget – SINR versus DL throughput




Link budget – RSRP versus UL throughput




Link budget – SINR versus UL throughput




Annex: eNB automatic neighbors list creation 1/1

Extract of a commissioning file after the eNB populated it’s cell neighbor list.



During commissioning of the eNB, only the IP address of the neighbor eNB is required. After the eNB has setup the X2 interface with the neighbor eNB, then it will exchange all necessary information of neighbor cells. That information is stored in the eNB commissioning file (see next slide)


Annex: eNB automatic neighbors list creation 2/2 Extract of a commissioning file after the eNB populated it’s cell neighbor list.




LTE Trial Test Result (Etisalat) -V1.6

Recommend Documents