High IRAT Failure in preparation phase

[Spider]

Hi experts,

In our E******* UTRAN we noticed a high preparation failure rate due mainly to the cause "RelocationFailureInTargetCNRNCorTargetSystem". I think that this is related to some capacity issues on 2G side, but from several flowcharts i saw the cause "Failure in the radio interface" sent in the "relocation cancel" message sent from the RNC, any idea about this??

Thanks

[gprastomo]

Yes it can be due to capacity issue or inconsistency, LAC,CI,BCCH BSIC etc
Check the inconsistencies then if its ok, then check the COngestion on TCH 2G

[dabreulima]

For everyone with the same problem, check the parameter Inter-RAT In BSC Handover Enable for 2G cells.
If this parameter is set = NO, will fail during the preparation of IRATHO to those cells.

[Spider]

Actually i've got the failure per cell couple (3G/2G), i think if the parameter Inter-RAT In BSC Handover Enable for 2G cells will be set to "NO" we will have 100% failures which is not our case, since failure rate per cell couple is <>100%. Actually we detect 2G congestion, can anyone suggest some parameter setting instead of 2G capacity extension?

[Spider]

Actually i got a proposition to change the parameters usedFreqThresh2dEcno to -14 and usedFreqThresh2dRscp to -108 to change the compress mode trigger, but i'm not convinced about this proposition since the 3G neighbours should also be optimized. Any comments about this?

[ninjafine]

Actually i got a proposition to change the parameters usedFreqThresh2dEcno to -14 and usedFreqThresh2dRscp to -108 to change the compress mode trigger, but i'm not convinced about this proposition since the 3G neighbours should also be optimized. Any comments about this?

Hi! I would first switch off all redirections to GSM (DR, LBHO, SBHO). Check also if all 2G->3G neighbours are defined properly. You may also enable HO from 2G to 3G. Then i would try to increase 3G coverage (PCPICH power, tilts).
Thanks and reps are appreciated.

[Spider]

I checked again and remark that the RNC is sending many Relocation required to the same cell and all are failed. Consequently we have a high preparation failure rate since the procedure is triggered many times. Anyone who knows the set of parameters in E// RNC which are involved in this?