IRAT Performance Improvement Tasikm ala alaya ya Re Reg g ion
www.huawei.com
Huawei Performance Team
Problem Statement and Route Cause Analysis Problem Statement: IRATHO CS and IRATHO PS SR performance degraded in Tasikmalaya (RNC Sukamenak). Major Problem was with IRATHO PS SR. Route Cause Analysis: In first level analysis we found that most of the IRAT failures (CS IRAT and PS IRAT both) were due to high physical channel failure at worst cells. Further to this for PS IRAT failure another reason was high failure due to PS UTRAN No Reply in poor RF conditions. After finding out 2 major reasons for CS and PS IRAT failures we investigate further and found bellow mentioned conclusions –
1. We started analysis for 3G to 2G neighbor level IRATHO performance for both CS and PS services. In this analysis we found that few target 2G cells have high congestion for CS and PS services, which was one of cause for IRAT failure due to physical channel failure. 2. IRAT failure due to PS UTRAN No reply was mainly caused by poor 2G RF conditions at time of handover.
Top Worst Cells
•
Parameter optimization done in 3 steps as per action plan mentioned in previous slide .(first on 13 th Sep , second on 17 th Sep and 3rd on 18th Sep 2012). After these actions IRAT performance improved and now maintained above baseline.
Handling Process : Handling Process: Now we know that route cause of poor IRAT performance was congestion at target 2G cells and poor 2G coverage at time of IRAT handovers. Capacity augmentation done by 2G team on request for congested 2G cells on 12 th Sep 2012 (Most affected 2G cell was CIANDA 1) and PS IRAT performance improved after this. We also done bellow mentioned parameter optimization to further improve IRAT performance as it was still bellow baseline 1. As per default setting in 3G side, we consider when the 2G signal is more than -95dbm, and the 3G signal meet the IRAT HHO condition, the IRAT HHO will happen, as we know, when the 3G signal is bad, 2G will also not very good, and 95dbm for 2G is not a good signal, so we planned to increase the 2G signal threshold and Inter-RAT HO Physical Channel Failure Penalty Timer to make sure when 3G want to handover to 2G, the 2G signal is good. we changed UCELLINTERRATHOCOV parameters TargetRatCsThd=16 to 26, TargetRatR99PsThd=16 to 26, TargetRatHThd=16 to 26, PenaltyTimeForPhyChFail=30 to 60 at worst cells. This change also implemented in two steps first on 13 th Sep and then on17th Sep 2012 at remaining worst cells. 2. We Planned to change U2GNCELL parameter Cell individual offset (CIOOffset) for the GSM cell. This parameter is used for decision making for the inter-RAT handover. A larger value of the parameter indicates the higher the handover priority of the GSM cell. A smaller the value of the parameter indicates the lower the handover priority of the GSM cell. So we reduced the priority of congested 2G cells for IRAT handover based on level of congestion and number of physical channel failure during IRAT. This action implemented on 17 th Sep 2012 in recent. 3. Now we decided to implement Action 1 for remaining worst cells of PS IRAT and also changed UEXT2GCELL Parameter RATCELLTYPE from EDGE to GPRS for those neighbors where high PS IRAT failure were occurring in poor coverage. This change done on 17 th Sep 2012. 4. After Action 2 and 3 PS IRAT Performance also improved but PS IRATHO SR was still slightly bellow baseline. So we finally decided to change UCONNMODETIMER Parameter T309 on 18th Sep 2012 in RNC Sukamenak. Timer T309 is started after the UE is reselected to a cell belonging to another radio access system in connected mode, or the CELL CHANGE ORDER FROM UTRAN message is received. It is stopped after the UE is successfully connected in the new cell. The UE will continue the connection to UTRAN upon expiry. Protocol default value is 5s. We want to allow some more time to UE to successfully connect to new cell. So W e changed T309 to 8s. And after changing this IRAT PS performance also improved and maintained above baseline. –
Details of Parameter Changes: Bellow mentioned UCELLINTERRATHOCOV parameters changed on 13 th and 17 th Sep 2012 .
Details of Parameter Changes: Bellow mentioned U2GNCELL Parameters changed on 17 th Sep 2012 .
Bellow mentioned UCONNMODETIMER Parameters changed on 18 th Sep 2012.
KPI Trend First Step
•
•
•
First Step
Second Step
Third Step
2G Data capacity augmentation done on 12th Sep 2012. Worst cell of 2G congestion was CIANDA 1. More PDCH channel added at other 2G cells also which have smaller data congestion in order to provide more room for data. PS IRAT performance improved after that. Parameter Optimization at 3G end done in 3 steps first on 13 th Sep , second on 17 th Sep and third on 18th Sep 2012. For details of actions in each step please refer previous slide named Handling process and details of parameter changes. After these actions IRAT performance improved and now maintained above baseline.
