Abnormally high RRC Setup Failures due to Uu No reply in some Cells

[theoptimizer]

Hi,

We have very high RRC Fail due to Uu No reply on some cells . Below is the summary of the issue

1) From the Trace we found that RNC is sending RRC Connection Setup but not receiving RRC Setup Complete from UE and after timer expiry it release the RL and count it as a a Failure due to No Reply.

RTWP is normal, I Changed Cell reselection thd and put hysteresis and modify Ssearchrat to make sure no ping pong happen between 2G & 3G . Cell Avg EcNo and RSCP is excellent. Still no improvement observed and no alarms on cell.

Did anyone have similar issue. Please support if you can

For some cells RRC for service is perfect issue if with only RRC others (Detach/IRAT Cell reselection/Registration) and for some cells all RRC is affected

Waiting for some ideas to troubleshoot this

[qaqa]

Based on your description the best way would be to do a drive test in the area to see if the UE is reselecting to other RATs or going out of service.
what is your QQUALmin and QRxlevmin?
to mr it looks like the UE is reselecting to GSM while in RRC connection process,

[theoptimizer]

Just to update i think we found the issue all cells suffering from this issue is having External Clock Problem Alarm. Does anyone know how to fix this using LMT ?

[rumy]

Did you solve the issue?
I presume the failures degraded your RRC SR; is it new 3G network?

[deknoi55]

Check TP first and how about physical area.

[npo.david]

For my experience:
Drive test or use Propagation Delay to detect any cell overshoot to that cell dominant area. Average EcNo and Rscp may good but where we have failure quality may not good.

Most effective for these cases:
- max transmit power of FACH, increase to 20
- constant value for calculating initial TX power, increase to -15

[brianm]

Just to update i think we found the issue all cells suffering from this issue is having External Clock Problem Alarm. Does anyone know how to fix this using LMT ?

Is the TX of the site ATM or IP? Make sure the CLK Mode is set correctly. For ATM it will be Line clock and for IP it will be SynchEth

However if its set correctly then the TX guys need to check if the clocking is set correctly from clock source (E1 or TX nodes)

[theoptimizer]

Hi,

The issue is still there but i find some more findings:

1) Only Registration and IRAT reslections RRC are having this issue
2) From CHR i found only UE category 7 & 8 are failing and i think all of them because the failures are in thousands like 30000 no reply fails on a cell per day

How come only some specific releases are suffering from this issue ? have anyone faced this kind of issue before ?

This is clearly abnormal

Some RNCs are not having this issue but from MML compare i didnt find anything mismatching that can lead to such abnormality

Please support

[rumy]

Check if you have SIM issues...it can be that the SIM is not allowed to 'roam' so many of the attempts are unsuccessfully due to old SIMs. Do you know if the operator has provided USIMs or has upgraded the profile in CN??? Your issue is leading toward this Core issue if the unsuccessful attempts are coming with the cause of Reg or Resel.

[theoptimizer]

Finally Solved!!!! its was the EFACH feature. We turned it off from UCELLALGO and no reply cleared. although it doesnt make sense and for sure its a bug/Issue in SW or feature. Waiting now for Huawei responce on this