PDA

View Full Version : Spurious rach



essalvador
2009-12-25, 06:54 PM
Bonjour,

Dans le matériel B9 d'alcatel et lors d'assignement SDCCH, je remarque toujours une congestion pour certains cellules pourtant leurs SDCCH_load_BH étant inférieur à 10% (leurs location update étant aussi inférieur à 50%).

J'ai examiné les voisines de ces cellules (pour pouvoir dire s'il y a des voisines qui ont les même couples bsic,bcch) mais toujours rien. En revanche, ils y a certains voisines qui ont les mêmes Bcch. Donc ma question est ce que avoir le même bcch d'une voisine et son homologue (son voisine du voisine) pourra générer un spurious rach (et donc une congestion SDCCH).

Si oui dois-t-on alors changer les BCCH ?

Merci d'avance!


En anglais:

Hello,

In hardware alcatel B9 at SDCCH assignment, I always notice some congestion but their SDCCH_load_BH cells was less than 10% (location update is also less than 50%).

I have reviewed the neighboring cells (so you can tell if there are neighbors who have the same couple BSIC, BCCH) but still nothing. However, they are some neighbors who have the same BCCH. So my question is what have the same BCCH of a neighbor and its counterpart (its neighbor's neighbor) will generate a spurious rach (and therefore congestion SDCCH).

If yes do we then change the BCCH?

Thank you in advance!

punto
2009-12-25, 11:14 PM
En anglais: ;)

Hi ... double BCCHs/BSICs pairs will cause you SDCCH assignment failure, of course , and also some problems with hardware on problematic cell, can cause this too. Repeated RACH , and SDCCH assignment failures (Fail-Radio) will cause congestion SDCCH, but not due to border LA update, SMS ... like you checked, already. I am sure ...
Non-GSM interference causes could be from other mobile networks TACS/AMPS ..., also.
Possible is also to have intermodulation with GSM BS/MS receiver, spurious RACH for AMPS - its Tx band is close to GSM UL band , and also radar, anti-theft mechanisms, medical devices ... can influence, too, slightly but threat exists.
This is tricky problem, anyway ... do make a BSC alerter for beginning to track where is happening mostly this problem - by network "nodes", affected ... set alerter to start if the SDCCH_assignment_failure >25% and to clear it at SDCCH_assignment_failure <15%. Of course you can play with this thresholds how you like ... but tell us if you found out some more about this ghost/phantom/spurious RACH ...

Pls, do try to shrink the cell - reduce on one test cell param RxLEvAccessMin for -3 dB to see how it will influence the SDCCH assign failure ... Best of luck, P:)

s52d
2010-10-01, 02:05 AM
Hi!

Beside what punto explained, you might get RACH failures by receiving
RACH to another cell (or, more likely, handover access bursts if TS0 is not alligned).
It is BCC part of BSIC to watch: BCCH 73 with BSIC 10 and BSIC 50 are considered equall.
More accuretely, training sequence (TSC) which is usually same as BCC.

To prevent it, we plan for BCC part, and on top of it we add NCC to separate cells even more.

BR
s52d