eRAB fail in one cell due to MME Packet loss in GBR

Hi All,
I have eRAB fail in one cell due to MME end Packet loss in GBR.
Anything can I do for parameter level in LTE?
Erab and rrc is 97 and issue is no context at mme.
Vendor is Huawei.

How to increase cell level erab and rrc?
Can I delay reselection to improve rrc and erab fail? As it will increase attempts?
How to reduce packet loss in that cell?
No rssi issue.

No context means that after HO, new eNB hasn’t gotten your context or wasn’t transfered via X2 or by MME.
Check X2 alarms, configure nbr etc.

But my rrc is failing not HO and erab.
I will check X2.
HO is also 97.
Maybe need to check.
If X2 failures, then I need to resolve that first.
But what about packet loss?
X2 can impact as it goes through s1?
Thanks for the suggestion.

My issue is a bit different…
I have LTE E-RAB failures due to MME in certain cells in one area.
Percentage of failures range from 2% to 10%.
All sites are connected to one hub from TX side.
Is this solely Transmission issue?

Can you check erab fail due to TNL?

Not TNL failures at all.

Do some sites ping test for confirmation, through OSS.

Super. Thanks for the tip!

What about coverage?
Are they remote looking cells?
Secondly, issue persist on few cells or on all cells?
Did you compare them with network or area average?

if there transmission issue ,MME delaying the NAS sending the initial UE context Setup message.
check eNB’s history alarm if there intermittent transmission alarm both on S1and X2 & check reason of failure from board log.

1 Like

Further checking could result in a coverage issue (overshooting), but I am wondering why the failure cause wasn’t No Reply.

In some networks T300 value is high.

Signaling phase somehow pass the RRC phase but unable to pass the ERAB Phase.
For example when a mobile try to attach they pass the RRC phase but fail the ERAB Phase due to say poor coverage.
Normally the MME timers are larger than UE timers. UE timer expire and UE re-initiate the attach procedure.

When UE reach the ERAB setup phase MME came to know that context of same UE is already present.
So MME discard the old request/context and treat the new one.
This type of failures pegged in MME releases.
It was due to one or few UEs in bad coverage but they degrade the ERAB setup success rate at cell level.

If you have CTR please do check the same. I found many such cases in Huawei.
It’s due to whether ERAB message size is large or anything related with the robustness, I am not sure.

But this phenomenon exists. Sometimes changing weather like rains, etc. Or change in indoor traffic profile e.g. lock down due to Covid also adds to the misery.

Pass or Establish. Whatever word seems appropriate to you :slight_smile:
If your vendor is Ericsson do check UL SINR, or Poor Coverage counters as well.

Final Addition: UE timer expires while MME timer is still running.
UE re-initiate the attach procedure and during this when UE reach the ERAB setup phase MME know that context of this UE is already present.
So MME discard the old request and treat the new one.

Many thanks for useful insights and descriptions! :wink:
Much appreciated!