3G to 2G IRAT Handover T-Mobile FSC UMTS Performance Team
1
AGENDA
T-Mobile IRAT Policy IRAT Handover Scope IRAT Neighbor List Creation IRAT Neighbor list planning and verification tools Ericsson 3G to 2G Handover Algorithm Nokia 3G to 2G Handover Algorithm IRAT Scenarios IRAT Parameter Recommendations Recommendations
2
AGENDA
T-Mobile IRAT Policy IRAT Handover Scope IRAT Neighbor List Creation IRAT Neighbor list planning and verification tools Ericsson 3G to 2G Handover Algorithm Nokia 3G to 2G Handover Algorithm IRAT Scenarios IRAT Parameter Recommendations Recommendations
2
T-Mobile IRAT Policy
IRAT should be used mainly for rescue mode handovers and not for traffic steering due to its longer interruption times and higher failure rates. During the initial network launch, CS traffic should be evenly divided between GSM and UMTS networks while PS traffic should be mainly directed to the UMTS network. When the PS traffic volume increases and becomes dominant, PS traffic should be directed to the UMTS network and CS traffic should be directed to the GSM network. UE should be preferably cell-camping on UMTS because of its superior data capability. 3
Optimal IRAT Policy Cell camping
CS Voice handover
PS Data handover/cell change
Load and or PS data traffic volume is low (< 20%) UE should camp on UMTS when first powered on UE should camp on UMTS with "aggressive" cell-reselection thresholds The UMTS cell-camping area allows to be expanded because the load is low. Handover should be used mainly for rescue reasons Load-based handovers may be used as a near term means to steer voice traffic to some degree but significant compromise in efficiency and quality can be expected. Handover mechanism of the following trigger-type should be supported. These handover mechanism should support both from UMTS to GSM and from GSM to UMTS: coverage -based quality-based load based service based The cell change and PS HO should be supported as detailed in the 3GPP inter-system Mobility Mechanisms
Load and or PS data traffic volume is high (> 20%) UE should camp on UMTS when powered on UE should camp on UMTS with "fair" cell-reselection thresholds. The UMTS cell-camping area should be set back to normal when the load is high
Handover should be used mainly for rescue reasons Load-based handovers may be used as a near term means to steer voice traffic to some degree but significant compromise in efficiency and quality can be expected. Handover mechanism of the following trigger-type should be supported. These handover mechanism should support both from UMTS to GSM and from GSM to UMTS: coverage -based quality-based load based service based The cell change and PS HO s hould be supported as detailed in the 3GPP inter-system Mobility Mechanisms
4
IRAT Handover Scope
Offload the GSM speech traffic by utilizing UMTS networks where available to carry CS traffic. Keep UMTS capable UEs on UMTS coverage as long as possible with an acceptable call quality.
• Handover 3G UEs to GSM while leaving UMTS coverage areas before call quality degradation. • 2G to 3G reselection can be used for 3G UEs to select a UMTS cell when back in UMTS coverage (reserved for future discussion). discussion). 5
3G to 2G Neighbor List Creation STRATEGY •
Create GSM neighbor lists for all UMTS cells in the network irrespective of their location.
•
Optimize the GSM neighbor list in order to minimize the time taken to make compressed mode measurements by UE.
•
Utilize existing 2G neighbor lists where possible to create 3G to 2G neighbor pairs.
•
Use planning tools for validation checks on 2G neighbor lists created from above step.
•
3 different solutions based on UMTS overlay configuration.
6
3G to 2G Neighbor List Creation Scenario 1 - 1:1 overlay with matched azimuths
UMTS site overlay with GSM site. UMTS azimuths are within +/- 15 degrees of GSM azimuths. UMTS uses antennas with a horizontal beam width equal to 65 or 90 degrees. Recommended size is 11 neighbors with 8 from the following procedure and 3 reserved for missing GSM cells after validation checks. 7
3G to 2G Neighbor List Creation Scenario 1 - 1:1 overlay with matched azimuths
Identify the underlying GSM cell for the UMTS cell. Add the underlying GSM cell to the top of the neighbor list. Add the other 2 sectors of the co-located GSM site to this list with priority 2 and 3 .
Using GSM statistics, order the remaining 2G neighbors of the GSM cell as per number of handover completions. Choose the top 5 neighbors and add it to the bottom of the above created list Validate the list using visual checks or planning tools to make sure none of the obvious neighbors are missing. Make sure that all required in-building GSM cells are included, otherwise add them as necessary. Extend the number of neighbors beyond 11 only if necessary to accommodate any missing 2G neighbors. 8
3G to 2G Neighbor List Creation Scenario 2 - 1:1 overlay with mis-matched azimuths
UMTS site overlay with GSM site. UMTS azimuths are more than +/- 15 degrees different from GSM azimuths. This solution can also be used for overlay sites with matched azimuths that have a narrow horizontal beam width. Recommended size is 14 neighbors with 11 neighbors from the following procedure and 3 reserved for any missing neighbors after validation checks.
9
3G to 2G Neighbor List Creation Scenario 2 - 1:1 overlay with mis-matched azimuths
Add all 3 sectors of the co-located GSM site to the top of the 3G -> 2G neighbor list. Identify the two closest co-located GSM sectors for the UMTS cell. Using GSM statistics, order the combined 2G neighbor list of these 2 GSM sectors as per number of handover completions. Choose the top 8 neighbors and add it to the bottom of the above created list. Validate the list using visual checks or planning tools to make sure none of the obvious neighbors are missing. Make sure that all required in-building GSM cells are included, otherwise add them as necessary. Extend the number of neighbors beyond 14 only if necessary to accommodate any missing 2G neighbors. 10
3G to 2G Neighbor List Creation Scenario 3 – UMTS site without overlay
UMTS site without a co-located GSM site. Create neighbor list from scratch using visual checks or planning tools as for a new site. Recommended size is 11 neighbors. Extend the number of neighbors beyond 11 only if necessary to accommodate any missing 2G neighbors. 11
3G to 2G Neighbor List Planning and Verification Tools The FSC RF Engineering team has created some tools that helps the engineers to check statistics, parameter settings, co-located cells, missing GSM neighbors and more. Under the GSM area the following features can be used to collect and plan GSM neighbors for UMTS cells: – New Site Performance Tool – Cell Configuration
The GSM neighbor list existing auditing features for UMTS cells are: – Co-located UTRAN and GSM cells – Missing co-located GSM neighbors – Missing GSM neighbors (coming soon)
The following slides present some snapshots of the tool. This tool can be accessed at the following link:
12
FSC RF Engineering Tool Reports Menu The following picture shows the FSC RF Engineering tool Reports Menu window.
13
FSC RF Engineering Tool Cell Configuration feature The cell configuration option allows the user to check the configuration and defined neighbors of the GSM cell.
14
FSC RF Engineering Tool New Site Performance Tool feature This feature provides statistical information for the cell performance. It also has a specific section for handover attempts that can help to prioritize the neighbors
15
FSC RF Engineering Tool Co-located UTRAN and GSM cells feature The cell configuration option allows the user to check which UMTS cells are colocated on GSM cells and what cell is that.
16
FSC RF Engineering Tool Missing co-located GSM neighbors feature This feature allows the engineer to check if there are any missing co-located cells in the GSM neighbors that have been defined for the UMTS cell.
17
Ericsson 3G to 2G Handover Algorithm
18
IRAT Feature Activation Parameters
C_GsmHoAllowed: IRAT handover is
only attempted if this parameter is set to Allowed for the current UeRc state. This is a constant hard coded value. FddGsmHoSupp : This should be set to TRUE to enable IRAT functionality.
19
IRAT Interaction with Inter Frequency Handover hoType: set per Cell. Indicates if IRAT Handover or Inter-Frequency HO or None shall be evaluated in case both Inter RAT and Inter-Frequency neighboring cells have been configured.
• • •
If all cells in the Active Set have hoType = None, neither Inter-Frequency nor InterRAT handover will be attempted. If at least one cell in Active Set has hoType = GSM-Preferred and no cell has hoType = IF-Preferred, IRAT handover will be attempted. If at least one cell in Active Set have hoType = IF-Preferred, Inter-Frequency handover will be attempted.
defaultHoType: Parameter set per carrier per DRNC. Indicates hoType for external cells. If not defined then IRAT is preferred. Recommendation: • Set hotype = GSM-Preferred for all cells at the edge of the network and around coverage holes. • When we have more than 1 carrier set hotype = GSM-Preferred for UMTS cells at the edge of coverage and set hoType = IF-Preferred for other cells within the core. This strategy can be reviewed when we have more than 1 carri er.
20
3G to 2G Handover Procedure
• Event criteria is triggered for connection quality monitoring based on – CPICH EcNo in DL – CPICH RSCP in DL – UE TX power in UL
• Event based GSM Measurements reporting • Identification of target cell • Inter-System HO execution 21
Event based connection quality monitoring • • •
•
• • • •
Connection quality is monitored on DL using CPICH RSCP and CPICH Ec/No Connection quality is monitored on UL using UE Tx power. Connection quality is checked in the following priority order: first downlink CPICH Ec/No, then downlink CPICH RSCP, then uplink Ue TX power. The measurement quantity is then set automatically depending on the check result. If at least one of the conditions (Rscp, Ec/No or UE Tx power) is bad, then RNC starts an IRAT handover attempt (if hoType = GSM-Preferred). If all the 3 conditions become good, then the IRAT handover attempt will be aborted. The DL Ec/No measurement can be disabled by setting TimeToTrigger2dEcno = 5000. The DL Rscp measurement can be disabled by setting TimeToTrigger2dRscp = 5000. The UL UE Tx measurement can be disabled by setting txPowerConnQualMonEnabled = False.
22
Concept of Filtering, Weighting and Offsets • •
The measured values are filtered by the UE before comparison with the event criteria. The filtering is performed according to the following formula: Fn = (1 –a) Fn-1 + a Mn Fn
is the updated filtered measurement result is the old filtered measurement result Mn is the latest received measurement result from physical layer m easurements a = ½^(k/2), where k is the parameter received in the IE "Filter coefficient". W hen K=1, a=0.707 and k=2, a=0.5 Fn-1
• •
A weighting factor is used to include active set cells other than the best cell in evaluation criteria for reporting events. The measured value after weighting will be N W 10 Log M i (1 W ) 10 LogM Best i 1 A
Mi is a m easurement result of a cell in the active set NA is the number of cells in the current active set MBest is the measurement result of the cell in the active set with the highest measurement result W is the weighting factor
Further detail on the weighting can be found in 3GPP TS 25.331: "RRC Protocol Specification".
•
•
•
A weighting factor of 0 (FSC recommended value) enables the evaluation based only on the best cell in the active set. By applying an offset, the UE sends a MEASUREMENT REPORT message back to the SRNC as if the measured value is offset higher (positive offset) or offset lower (negative offset) than reality. The parameter individualOffset can be used for this purpose. Improper use of non-zero offset value may result in instability and unequal cell borders. FSC recommended value for individualOffset = 0. 23
Concept of Event 2d and 2f reporting
24
Event 2d/2f: CPICH EcNo in DL •
Event 2d is used to evaluate if a connection quality is bad.
•
Event 2f is used to evaluate if the connection quality is good.
usedFreqThresh2dEcno: indicates threshold to trigger event 2d for Ec/No. hysteresis2d : Hysteresis used for event 2d. timeToTrigger2dEcno : Interval time between detection of event 2d and sending of the measurement report, when the measurement quantity is Ec/N o. usedFreqRelThresh2fEcno :Relative threshold for event 2f for Ec/No. hysteresis2f : Hysteresis used for event 2f. timeToTrigger2fEcno: Interval time between detection of event 2f and sending of the measurement report, when the measurement quantity is Ec/No filtercoefficient2 : coefficient used to filter measured values before comparing to event criteria. usedFreqW2d: weighting factor for event 2d. usedFreqW2f : weighting factor for event 2f.
25
Event 2d/2f: CPICH RSCP in DL •
Event 2d is used to evaluate if a connection quality is bad.
•
Event 2f is used to evaluate if the connection quality is good.
usedFreqThresh2dRscp: indicates threshold to trigger event 2d for RSCP. hysteresis2d : Hysteresis used for event 2d. timeToTrigger2dRscp : Interval time between detection of event 2d and sending of the measurement report, when the measurement quantity is RSCP. usedFreqRelThresh2fRscp :Relative threshold for event 2f for RSCP. hysteresis2f : Hysteresis used for event 2f. timeToTrigger2fRscp: Interval time between detection of event 2f and sending of the measurement report, when the measurement quantity is RSCP. filtercoefficient2 : coefficient used to filter measured values before comparing to event criteria. usedFreqW2d: weighting factor for event 2d. usedFreqW2f : weighting factor for event 2f
26
Event 6d/6b: UE Tx power exceeds maximum value •
Event 6d is used to evaluate if a connection quality is bad.
•
Event 6b is used to evaluate if the connection quality is good.
•
There is no threshold to trigger event 6d. The UE Tx power is compared to its maximum transmit power P_max in order to trigger IRAT measurements.
txPowerConnQualMonEnabled: Enables or disables the connection quality monitoring based on UE Tx power. ueTxPowerThresh6b: The threshold used for trigger event 6b when UE TX power becomes less than an absolute threshold. timeToTrigger6d: Interval time between detection of event 6d and sending of the measurement report. timeTrigg6b: Interval time between detection of event 6b and sending of the measurement report. filterCoeff6 : Coefficient for layer 3 filtering before UE internal measurement reporting evaluation.
27
Active Set Update Effect on connection quality monitoring •
The parameters hoType, usedFreqThresh2dEcno and usedFreqThresh2dRscp are the only cell based IRAT parameters.
• All other IRAT related parameters are RNC-based and will have the same value for all cells. •
If the value of 2d threshold changes as a result of an Active Set update for the Ec/No or the Rscp measurement, due to different settings of the cell parameter usedFreqThresh2dEcno or usedFreqThresh2dRscp, the corresponding measurements using event 2d and event 2f must be modified with the new thresholds.
28
Event 3a - GSM Measurement Reporting • • • • •
When the connection quality is bad, measurements on GSM cells are act ivated. The UE is configured to report measurements according to event 3a. Ongoing IRAT measurements might need to be modified as a result of an Active Set update. If the GSM monitored set is changed, the new GSM monitored set must be sent to the Ue When an IRAT handover attempted is started, the same measurement quantity that triggered the bad quality is used. If connection quality trigger is CPICH Ec/No event 3a is triggered when the estimated quality < usedFreqThresh2dEcno + utranRelThresh3aEcno - hysteresis3a/2 and measured GSM carrier RSSI > gsmThresh3a, for at least TimeToTrigger3a CPICH RSCP event 3a is triggered when the estimated quality < usedFreqThresh2dRscp + utranRelThresh3aRscp - hysteresis3a/2 and measured GSM carrier RSSI > gsmThresh3a, for at least TimeToTrigger3a UE Tx power event 3a is triggered when the estimated quality < usedFreqThresh2dRscp + utranRelThresh3aRscp + utranRelThreshRscp - hysteresis3a/2 and measured GSM carrier RSSI > gsmThresh3a, for at least TimeToTrigger3a
29
Concept of Event 3a reporting
30
Event 3a – Parameter definitions utranRelThresh3aEcno: Relative threshold for event 3a versus event 2d when the 2d measurement quantity is CPICH Ec/No. utranRelThresh3aRscp: Relative threshold for event 3a versus event 2d when the 2d measurement quantity is CPICH RSCP. timeToTrigger3a: Time between detection of event 3a and sending of the measurement report. Hysteresis3a: Hysteresis used for event 3a. gsmThresh3a: Threshold for event 3a (the estimated quality of the currently used UTRAN RAN frequency is below a certain threshold and the estimated quality of the GSM system is above a certain threshold) for GSM. individual offset: The offset is added to the measured quantity before the UE evaluates whether an event has occurred. Improper use of non-default values may result in instability and unequal cell borders. Recommended value is 0 for network launch. maxGsmMonSubset: Maximum number of GSM cells that the UE will measure on. It is recommended to keep this value to the default = 32. utranFilterCoefficient3: Coefficient for layer 3 filtering of UTRAN quality before inter-RAT reporting evaluation. utranW3a: Weighting factor for event 3a for UTRAN.
31
GSM Measurements Cancellation •
When the UE is measuring for IRAT Handover, the evaluation is stopped in case – the UeRC state does not allow the evaluation to continue (e.g. due to rate connection switching) – the GSM Monitored Set becomes empty at Active Set update – the connection quality becomes good, i.e. if an event 2f or event 6b or both occur before event 3a occurs.
32
Identification of Target Cell •
If the RNC receives a second event 3a MEASUREMENT REPORT while it is already processing one, the new report is buffered.
•
RNC uses the measurement report from event 3a and orders the GSM cells according to their GSM quality measure.
•
It discards all cells whose GSM Carrier RSSI is below GsmThresh3a, and cells that are not BSIC verified.
•
RNC chooses the GSM cell with the best RSSI as the target cell for IRAT handover.
•
If the IRAT handover to the target cell fails, and if there are soft handover reports buffered, RNC uses these to perform soft handover.
•
If the IRAT handover fails, and there is a buffered 3a measurement report, RNC uses this buffered report to find a new GSM target for handover.
33
Identification of Target Cell Contd. •
If the IRAT handover fails with cause "Relocation Preparation Failure" or “Handover/CellChange from Utran Failure” and there is more than one GSM cell in the 3a measurement report, repeated attempts shall be made for these cells in quality order based on the following parameters.
gsmAmountPropRepeat: Maximum number of repeated attempts (not including the first attempt) of GSM cells for handover based on the same measurement report. gsmPropRepeatInterval: Minimum time interval between proposals of the same GSM cell for handover based on the same measurement report. •
If the IRAT handover failure cause is “Other”, then the IRAT algorithm will be terminated and the UE stays on UMTS.
34
Nokia 3G to 2G Handover Algorithm
35
IRAT Feature Activation Parameters •
GsmHandoverAMR: determines whether an IRAT handover to GSM is allowed for circuit-switched voice services.
•
GsmHandoverCS: determines whether an IRAT handover to GSM is allowed for circuit-switched data services.
•
GsmHandoverRtPS: determines whether an IRAT handover (cell change) to GSM/GPRS is allowed for real-time packet-switched data services in CELL_DCH state of connected mode
•
GsmHandoverNrtPS: determines whether an inter-system handover (cell change) to GSM/GPRS is allowed for non real-time packet-switched data services in CELL_DCH state of connected mode
36
IRAT Interaction with Inter Frequency Handover •
•
If a UMTS cell has both IRAT and Inter frequency neighbors, the RNC determines the priorities between inter-frequency and -system handovers on the basis of Service Handover IE value. The RNC receives the Service Handover IE from the core network in the RAB ASSIGNMENT REQUEST or RELOCATION REQUEST (RANAP) message. –
–
–
•
Should be handed over to GSM: Handover to GSM has priority over the inter-frequency handover. In this case the RNC shall not start inter-frequency measurements until the inter-system (GSM) measurements are completed. Should not be handed over to GSM: Inter-frequency handover has priority over the handover to GSM. In this case the RNC shall not start the GSM measurements until the inter- frequency measurements are completed. Shall not be handed over to GSM: Inter-frequency handover has priority over the handover to GSM. In this case the RNC shall not start GSM measurements or handover to GSM even if no neighboring cell is good enough for the quality and/or coverage reason inter-f requency handover. This means that the RNC does not initiate handover to GSM for the UE unless the RABs with this indication have first been released with the normal release procedures.
If the RNC does not receive the Service Handover IE from the core network, interfrequency handover has priority over inter-system handover as a default value.
37
3G to 2G Handover Procedure • Event criteria is triggered based on – CPICH EcNo in DL – CPICH RSCP in DL – UE TX power in UL
• Other available criteria to trigger IRAT measurements are – BTS Downlink radio link power exceeds threshold – UL dedicated traffic channel quality deterioration report triggered by Outer Loop Power Control Recommendation is to enable IRAT handover based on Cpich Ec/No, Cpich Rscp and UE Tx power for network launch.
• Periodic GSM Measurements reporting • Identification of target cell • Inter-System HO execution 38
Event based measurement triggering •
FSC recommendation is to limit IRAT measurements triggering to CPICH RSCP, CPICH Ec/No and UE Tx power for UMTS network launch.
•
Low CPICH RSCP or Low CPICH Ec/No compared to pre-defined thresholds trigger event 1F.
•
Event 1E is triggered when CPICH RSCP or CPICH Ec/No becomes better than the defined thresholds.
•
Event 6A is triggered when the UE Tx power becomes greater than a pre-defined threshold.
•
Event 6B is triggered when the UE Tx power decreases compared to the pre-defined threshold.
•
The handover cause, which has triggered first has the highest priority. RNC does not stop or modify ongoing GSM measurements and handover decision procedures if another handover cause is triggered during the handover procedures.
•
In case of low CPICH RSCP or low CPICH Ec/No, RNC starts GSM measurements only when event 1F is triggered for all cells in the active set.
•
The RNC does not break off ongoing inter-system measurements if it receives event 1E or event 6B triggered measurement reports from the UE during the IRAT Handover process
39
Concept of events 1F and 1E CM
NO
NO
NO
NO
NO YES
Cell A Cell B
OK OK
OK OK
1F OK
OK OK
1F OK
1F 1F
Cell C
OK
1F
1F
OK
1F
1F
Ec/No Cell A Cell B Cell C
HHOEcNoCancel HHOEcNoThreshold
HHOEcNoTimeHysteresis
HHOEcNoCancelTime
t
Cell C meets the threshold for event 1F but not for
Cell A triggers
Cell A triggers
event 1F
event 1F
HHOEcNoTimeHysteresis
Cell A triggers Cell C triggers
Cell B triggers
40
Parameter definitions for low measured CPICH EcNo in DL GSMcauseCPICHEcNo: indicates whether IRAT handover to GSM based on low measured Ec/No is enabled. HHOEcNoThreshold: indicates EcNo value to trigger event 1F. HHOEcNoTimeHysteresis: time period during which the CPICH Ec/No of the active set cell must stay worse than the threshold HHoEcNoThreshold before the UE can trigger reporting event 1F. HHOEcNoCancel: EcNo value to trigger event 1E. (HHOEcNoCancel> HHOEcNoThreshold) HHOEcNoCancelTime, determines the time period during which the CPICH Ec/No of the active set cell must stay better than the threshold HHoEcNoCancel before the UE can trigger reporting event 1E. EcNoFilterCoefficient, controls the higher layer filtering of physical layer CPICH EcNo measurements before the event evaluation and reporting is performed by the UE. 41
Parameter definitions for low measured CPICH RSCP in DL GSMcauseCPICHrscp, indicates if HO to GSM due to low CPICH RSCP is enabled. HHORscpThreshold, indicates RSCP value to trigger event 1F. HHORscpTimeHysteresis, time period during which the CPICH RSCP of the active set cell must stay worse than the threshold HHoRscpThreshold before the UE can trigger reporting event 1F. HHORscpCancel, RSCP value to trigger event 1E. (HHORscpCancel> HHORscpThreshold) HHORscpCancelTime, determines the time period during which the CPICH RSCP of the active set cell must stay better than the threshold HHoRscpCancel before the UE can trigger reporting event 1E. HHORscpFilterCoefficient, controls the higher layer filtering of physical layer CPICH RSCP measurements before the event evaluation and reporting is performed by the UE. 42
Parameter Definitions for high UE Tx power GSMcauseTxPwrUL: indicates if HO to GSM due to UE Tx power is enabled. HHoMaxAllowedBitrateUL: defines the bit rate threshold that the maximum allocated user bit rate in the uplink DPCH may not exceed, so that the GSM handover is possible due to high uplink DPCH power level. GsmUETxPwrThrAMR: UE TX power threshold for CS Voice connection. GsmUETxPwrThrCS: UE TX power threshold for CS data connection. GsmUETxPwrThrNrtPS: UE TX pwr threshold for a non real time PS data connection. GsmUETxPwrThrRtPS: UE TX pwr threshold for a real time PS data connection. GsmUETxPwrThrTimeHyst, determines the time period during which the UE transmission power must stay below the transmission power threshold before the UE calls off the handover cause (event 6B). GsmUETxPwrFilterCoeff , controls the higher layer filtering of physical layer UE Tx pwr measurements before the event evaluation and reporting is performed by the UE •
The power thresholds are relative to the max TX power level a UE can use on the DPCH in the cell (or the max RF output power capability of the UE, whichever is lower). 43
BTS DL radio link power exceeds threshold • • •
The base station (BTS) measures and averages the downlink code power of each radio link separately and reports the averaged measurement results to the controlling RNC. These measurement reports can trigger IRAT handover to GSM when the downlink transmission power of the radio link approaches its maximum allowed power level. The RNC starts measurements on GSM cells if the DPCH TX power of a single radio link exceeds the threshold.
GSMcauseTxPwrDL: indicates if HO to GSM due to high DL DPCH TX pwr is enabled HHoMaxAllowedBitrateDL: defines the bit rate threshold that the maximum allocated user bit rate on the downlink DPCH may not exceed, so that the GSM handover is possible due to high downlink DPCH power level GsmDLTxPwrThrAMR: DL DPCH TX pwr threshold for CS Voice connection. GsmDLTxPwrThrCS: DL DPCH TX pwr threshold for CS data connection. GsmDLTxPwrThrNrtPS: DL DPCH TX pwr threshold for a non real time PS data connection GsmDLTxPwrThrRtPS: DL DPCH TX pwr threshold for a real time PS data connection RLMeasRepPeriod: defines the reporting period of the Radio link measurements.
•
The DL DPCH TX power threshold is relative to the allocated maximum transmission power of the DPCH
44
UL Quality deterioration report triggered by Outer Loop Power Control •
• • •
The uplink outer loop power control sends the quality deterioration report to the handover control, if the uplink quality stays constantly worse than BLER target although the uplink SIR target has reached the maximum value (the UE has reached either its maximum Tx power capability or the maximum allowed transmission power level on the DPCH). This quality deterioration report from the uplink outer loop power control can be used to trigger off inter-system handover to GSM. The uplink outer loop PC repeats the quality deterioration reports to the handover control periodically until the uplink SIR target decreases below the maximum value. Handover control does not interrupt an ongoing inter-system (GSM) measurement procedure even if the uplink outer loop PC stops sending the quality deterioration reports.
GSMcauseUplinkQuality: indicates whether an IRAT handover to GSM caused by Uplink DCH quality is enabled. HHoMaxAllowedBitrateUL: defines the bit rate threshold that the maximum allocated user bit rate in the uplink DPCH may not exceed, so that the GSM handover is possible due to high uplink DPCH power level EnableULQualDetRep: indicates whether the uplink outer loop PC can send a quality deterioration report to handover control when the quality stays worse than the BER/BLER target despite of the maximum uplink SIR target. ULQualDetRepThreshold: the period during which the quality must constantly stay worse than the BER/BLER target before the uplink outer loop PC may send a quality deterioration report.
45
Periodic GSM Measurements Reporting • • •
UE makes measurements on GSM neighbors using compressed mode. The RNC makes the handover decision on the basis of the periodical inter -system measurement reports received from the UE. GSMMeasRepInterval defines the reporting interval for these reports. UE measures all possible GSM neighbors during GSMNcellSearchPeriod, during which no handover decision is made.
GSMMeasRepInterval: determines the measurement reporting interval for periodical GSM measurements. GSMNcellSearchPeriod: determines the number of periodical GSM measurement reports, starting from the first report after the measurement setup, during which a handover to GSM is not possible. This period allows the UE to find and report all potential GSM neighbor cells before the handover decision. GsmMaxMeasPeriod: determine the max allowed duration of the GSM measurement. If the RNC is not able to execute the handover to GSM within this time, it stops the GSM measurement. GsmMinMeasInterval: determines the min interval between an unsuccessful IRAT measurement or handover procedure, and the following GSM measurement procedure related to the same RRC connection. Repetitive GSM measurements are disabled when the value of the parameter is zero GsmMinHoInterval: determines the minimum interval between a successful handover from GSM to UTRAN and the following IRAT handover attempt back to GSM related t o the same RRC connection.
46
Concept of GSM Measurements Reporting
1. GSMMeasRepInterval 2. GSMNcellSearchPeriod 3. GsmMaxMeasPeriod 4. GsmMinMeasInterval
47
Identification of Target Cell • •
•
• • •
•
•
After the search period has expired, RNC evaluates the radio link properties of the best GSM neighbor cells after every measurement report. The RNC initiates a handover attempt to the best GSM neighbor (target) cell as soon as the best GSM neighbor cell satisfies the required radio l ink properties. If there are several GSM cells which satisfy the required radio link properties, the RNC ranks the potential GSM cells according to the priority levels and selects the highest ranked GSM cell to be the target cell. A cell is ranked higher than another cell if it has a higher priority level even though its signal strength condition is worse. Signal strength conditions have effect only between cells which have the same priority level. FSC Recommendation is to datafill all GSM neighbors for each UMTS cell with same priority so that the IRAT target is found based on signal strength conditions. Exception to this rule should be used only on a as required basis after network launch. In the case of CS, RNC requests BSIC verification before IRAT handover to enable the UE to synchronize on the target cell, as well as to identify the target when there is more than one GSM cell with the same BCCH frequency. In case of PS, RNC requests BSIC verification only when there is more than one GSM cell with same BCCH frequency.
48
Identification of Target Cell Contd •
The RNC initiates a handover attempt to the best GSM neighbor (target) cell as soon as the best GSM neighbor cell satisfies the required radio link properties. AVE_RXLEV_NCELL(n) > AdjgRxLevMinHO (n) + max( 0, AdjgTxPwrMaxTCH (n) - P_MAX ) – – –
AVE_RXLEV_NCELL(n) is the averaged GSM carrier RSSI value of the GSM neighbor cell (n) P_MAX indicates the maximum RF output power capability of the UE (dBm) in GSM AdjgTxPwrMaxTCH indicates the maximum transmission power (dBm) a UE may use in the GSM neighbor cell (n)
AdjgRxLevMinHO : minimum required RSSI (dBm) level which the averaged RSSI value of the GSM neighbor cell (n) must exceed before the inter-RAT handover is possible. AdjgTxPwrMaxTCH: the maximum transmission power (dBm) a UE may use in the GSM neighbor cell. GsmMeasAveWindow : maximum number of periodical GSM measurement reports (maximum size of the sliding averaging window) from which the RNC calculates the averaged GSM RSSI values for the handover decision algorithm. AdjgPriorityCoverage : The priority level of a GSM neighbor cell is controlled by this parameter.
49
IRAT Scenarios Based on mobile user behavior and their location in the UMTS network two cr iteria have been identified that can be used to define IRAT scenarios: •
Speed (Highway or Outdoor) This refers to how fast the UE is moving through the UMTS network. At these speeds the UE has minimum time to execute an IRAT handover once the UMTS coverage degrades. Therefore, the time to trigger the event has to be low.
•
User location (Network Core or Edge) This case corresponds to the physical location of the UE in the UMTS network. For users around the Edge of coverage, that have a higher probability of leaving the UMTS coverage, the RSCP and Ec/No thresholds to trigger GSM measurements can be set to higher levels to trigger the handover sooner. For those users in the core area, that are experiencing some degradation in the service it is more probable that their RF condition may recover. Thus, it is advisable to trigger the compressed mode a little later at lower levels of Ec/No and RSCP (without compromising the quality and retainability of the call) in order to maintain the call on UMTS as long as possible.
50
IRAT Scenarios Scenarios based on combination of both criteria • • • •
Highway at Network Edge Highway at Network Core Outdoor at Network Edge Outdoor at Network Core
T-Mobile USA will be deploying UMTS services using Ericsson and Nokia equipment. Each vendor provides different set of parameters and granularity. Therefore, not all scenarios may be applicable for both of them. The following summarizes the achievable scenarios for each vendor: Nokia IRAT scenarios: Highway at Network Edge Highway at Network Core Outdoor at Network Edge Outdoor at Network Core
Ericsson IRAT scenarios: Network Edge Network Core
51
IRAT Scenario Matrix
Highway Outdoor
Edge of network
Core of network
Less Time To Trigger
Less Time To Trigger
Higher Thresholds
Lower Thresholds
More Time To Trigger
More Time To trigger
Higher Thresholds
Lower Thresholds
52
IRAT Scenarios •
Highway at Network Edge This scenario refers to the case where a UE is leaving the UMTS coverage area at high speeds. At these speeds the UE has minimal time to execute an IRAT handover once trigger conditions are met. Thus, the time to trigger the event has to be low (fast). For users located in the network edge, it is advisable that compressed mode measurements are triggered sooner at higher RSCP and Ec/No levels, so as to give the UE enough time to perform the handover evaluation before a call is dropped.
•
Highway at Network Core This scenario corresponds to the case where a UE is moving at high speeds within the core of UMTS coverage. Due to the user high speed the UE should initiate GSM measurements without delay once the trigger conditions are met. Therefore, the time to trigger can be set to a low value, so that the threshold conditions will have to be met only for a short time. Because the UE is in the core of the UMTS network, it is advisable to trigger the compressed mode later at lower levels of Ec/No and RSCP (without compromising the quality and retainability of the call) in order t o maintain the call on UMTS as long as possible.
53
IRAT Scenarios •
Outdoor at Network Edge This scenario refers to the case where a UE is moving around the network edge with relatively low speeds. At low speeds the UE has some extra time to trigger measurements and execute an IRAT handover compared to the highway case. Hence, the same RSCP and Ec/No threshold for Highway at Network Edge can be used but allowing a longer time to start compressed mode. In this case more samples are considered for the trigger decision.
•
Outdoor at Network Core This scenario depicts the case where a UE is moving in the network core with relatively low speeds. At low speeds the UE has some extra time to trigger measurements and execute an IRAT handover compared to the highway case. While being in the network core, a call should be maintain as long as possible in UMTS without compromising its quality and retainability. Therefore, lower RSCP and Ec/No threshold can be used to trigger compressed mode a little later giving an opportunity to the UE to recover from a relative bad quality condition.
54
Nokia IRAT Scenarios – Parameter Settings •
•
•
•
Highway at Network Edge HHoEcNoThreshold = -12 dB HHoEcNoTimeHysteresis = 100 ms HHoRscpThreshold = -98 dBm HHoRscpTimeHysteresis = 100 ms
HHOEcNoCancel = -9 dB HHOEcNoCancelTime = 1280 ms HHORscpCancel = -95 dBm HHORscpCancelTime = 1280 ms
Highway at Network Core HHoEcNoThreshold = -14 dB HHoEcNoTimeHysteresis = 100 ms HHoRscpThreshold = -100 dBm HHoRscpTimeHysteresis = 100 ms
HHOEcNoCancel = -11 dB HHOEcNoCancelTime = 1280 ms HHORscpCancel = -97 dBm HHORscpCancelTime = 1280 ms
Outdoor at Network Edge HHoEcNoThreshold = -12 dB HHoEcNoTimeHysteresis = 320 ms HHoRscpThreshold = -98 dBm HHoRscpTimeHysteresis = 320 ms
HHOEcNoCancel = -9 dB HHOEcNoCancelTime = 1280 ms HHORscpCancel = -95 dBm HHORscpCancelTime = 1280 ms
Outdoor at Network Core HHoEcNoThreshold = -14 dB HHoEcNoTimeHysteresis = 320 ms HHoRscpThreshold = -100 dBm HHoRscpTimeHysteresis = 320 ms
HHOEcNoCancel = -11 dB HHOEcNoCancelTime = 1280 ms HHORscpCancel = -97 dBm HHORscpCancelTime = 1280 ms
Note: all these parameters are configurable at cell level and can be grouped in templates.
55