- This topic has 11 replies, 1 voice, and was last updated 3 years, 7 months ago by LEMAURE.
4th May 2011 at 07:46 #66281anhGuest
We’re facing low paging success rate (~83%) in one BSC (happen to all cells in the BSC)
We checked paging dimensioning:
– parameter settings: T3212 (equal to IMSI attach/detach timer at MSC), block for access grant: OK.
– SDCCH Congestion, MSC-BSC signalling links: OK.
– Paging load is low.
– BSC uses only 1 LAC.
From MSC, more than 10% of failure is due to “No response from MS”
The MSC is MSC-Server of Ericsson.
The BSC is Alcatel-Lucent B10.
What else can I check? Pls advice.
Thanks,4th May 2011 at 13:39 #66282SmartGuest
It happened since the installation or it started after your changes?5th May 2011 at 02:25 #66283anhGuest
It happened since the installation.5th May 2011 at 07:51 #66284SmartGuest
Maybe is a fraud case. Someone is using your card for VoIP. check the MS IMSI/IMEI if it is the same in every failure ban him.5th May 2011 at 14:49 #66285PixGuest
Basically the MSC is not able to page the MS. The unreachable MS are either off coverage or out of battery : those are the MS that could not detach properly from the network.
in order to avoid this “ghost” MS, have you tried decreasing the VLR purge timer ? try to set it to “2*T3212 + 10mn” or something like this.
pix13th May 2011 at 05:02 #66286ClintonAGuest
I note your BSC is ALU and your MSC is Ericsson, therefore I deduce there ‘maybe’ an issue with inconsistent parameter settings.
Typically in an Ericsson network, the T3212 parameter should be set relative to MSC parameter BTDM (Base Time Duration), which results in detach.
BTDM is the time supervision for implicit detach and it is crucial for BTDM to be higher than T3212 (i.e. BTDM ≥ T3212) so that the MS is not detach from the network before the period registration kicks in.
If BTDM is less than T3212 (BTDM < T3212), this will result in the MS getting detached earlier than MS doing periodic location update. The typical default values I have seen for T3212 is 30 or 40 (3 – 4 hours), with 40 being the normal default. In Ericsson the units of T3212 and BTDM (for Ericsson MSC) are different as well, i.e. T3212 uses deci-hour and BTDB uses minutes. I am unsure what the settings are in your ALU BSC, however please ensure your T3212 timer in ALU BSC is less than or equal to the BTDM in your Ericsson MSC.13th May 2011 at 08:59 #66287anhGuest
T3126=T+2S; T (Tx_integer)=32; S=217 RACH frames (~998.2ms)
So, I think the param. settings are fine. But, PSR is still low.
Thanks,14th May 2011 at 19:08 #66288pixGuest
yes, the timers are fine. I’m not familiar with the guard timer (GTDM) though.
It seems quite radical to have T3212 = BTDM. I would allow few more minutes for the BTDM. But anyway, that’s not the problem here.
so let’s focus somewhere else.
perhaps it is due to bad coverage in some areas.. or unstable BTS that provides coverage only once in a while.
before the T3212 expires, the MS is actually ALREADY out of coverage and cannot be reached. During this time, the MSC still has the Ms attached, but physically the Ms is not reachable.
however your timers are already quite short (108mn), which is 3 times less than standard duration. The benefit is that such “unreachable” situation might last ( in the worst case ) 107mn max.
in order to verify this hypothesis, you could try to decrease T3212 and BTDM. If you see an improvement, then… good.
also, do what Smart asked : trace the IMSIs which are unanswered. If they are always the same, then… that gives some pointers.
pix16th July 2011 at 11:08 #66289vndnguyenGuest
I am also facing with low paging success rate.
Related to your problem, in my opinion, T3101=3sec is a bit low and T3113=12sec is too high. I think you could increase T3101 to 4-5s and decrease T3113 to 9-10s.10th October 2012 at 11:57 #66290DipsyGuest
Configuration of Assocaited HLR on the MSC
Scenrio: having 3 HLR in Pool and .
an associated, to be seeing the HLRs as one on the MSC.
I have configured the M3UA link for the main HLR, am aslo using the links for the 3 HLR as the link for communication to the associated HLR, SCCPGT is configured and the SSN status is allowed
On the VLR request i can find the information of the sub on the MSC. What do u think is the problem?5th April 2017 at 17:23 #66291AyatGuest
We are also facing the same issue % psr % is loe 82 per LAC
I think it is related to radio problem but one of our engineer advice to use third paging , we only use first and 3 time
So do you think it will solve the problem , i am alawys arrange to check the timer and the relation is fine
Pix about purge16th April 2017 at 09:22 #66292LEMAUREGuest
– what about paging discard on LAPD link? (is your paging load per LAC>80%? if yes you may need to split concerned LACs or add a secondary broadcast channel on heavy cells)
– what is the value of your immediate assignement success rate per LAC as it also happen to reduce PSR? (congestion on signalling channel,…)
– check on the BCCH frequency plan and make sure interferences are minimised.