[Dinesh]

In most of the IDR(satelliet Media) sites in our N/w, it is very difficult to mature MT Calls. We have incorporated the delay also in the BTS. Even then the problem could not be sorted out. Please help

[pix]

dinesh,

are you basing your conclusion on qos stats or on drive tests ? which indicator is poor, exactly ?

regards

[Dinesh]

my observation is based on my practical experience.

i) I got many complains from those sites regarding the problem of Incoming Calls.

ii) I had personally tested and found that the problem is there. Whenever i call to someone at those sites, most of the time(60-70%) I got the “Not Reachable” announcement. But at that very moment they could call me without much difficulties.
NB- We use SIGDEL-LONG, TEI-40

[pix]

dinesh,

ok – and what is the error : number unreachable or network error ?

in terms of qos stats, can you see what’s wrong? is it sdcch assignment failures ? or paging failures?

i would say that either
(1) the called party is detached from the network (due to location update problems), or
(2) the paging phase is not completing properly
(3) the immediate assignment phase after the paging is not completing properly.

(3) would have a symmetrical impact on MO calls
about (1), it seems unlikely, but just to be sure, let’s do the following test : once a MS on this BTS has called you, can you call it back without any problem ?

if yes – then MS was indeed detached
if no – then MS attach/detach is not a problem.

about(2) – check your paging timer T3113 (as far as i remember), and there are perhaps others.

about (3) – here again, i would focus on the timers involved (there are more than 1)

Can you do a A interface trace? See what is happening after the paging is sent?

Regards
pix

[Dinesh]

Hi Pix,
Thanx for the advice. I think its the problem related to PAGING. Would you be kind enough to shed some light on various Paging Timers- Timers and their ideal values.

Also please let me know how to conduct A interface trace.

BR//

[pix]

hi dinesh,

well, what’s your value of T3113 ? if it’s about 3s, then that’s the default value. it should be enough, even with satellite delay. you could try to increase it to 5s (but i’m afraid it’s gonna impact your whole network !!)

i’m not familiar with other paging timers. i think you’ll have to check the 3GPP by yourself, or wait for others to answer. those timers are internal to the MSC. At this point, you should work with your colleagues at MSC side, they know those timers and they might know how to do an A interface trace.

to perform A interface trace, you need a protocol analyzer, such as tektronix K12 or K15.

regards
pix

[SHELDON]

Hi Dinesh,

Yes, I think your problem is with paging.

I would suggest you put those sites on a separate location area, because the delay in delivering the paging message to subs in these sites is high, and this will be exacerbated if these messages are put in a queue with pages for other MSs in a big LA, thus increasing the chance of failure.

Good luck.

SHELDON

[pix]

sheldon – Brilliant idea! I was wondering how to “exarcebate” such paging problems. Creating a dedicated LAC will do just that.

[Dinesh]

Thanx Pix & Thanx Sheldon,

I will try again by keeping those sites in a separate LAC.

Meanwhile I changed the TTALONG in my BSC to 15 from 25. There seem to be some improvement in the performance- althaugh yet to see in the QOS Report. Could you please elaborate the role of TTALONG and TTA in PAGING. What are the ideal values for the satellite sites.

Another thing I would like to Know about is the T3113 vale. How do we chk and manipulate the value. I tried using ALex Library but couldnt succeed. Pls Help.
BR//
Dinesh

[pix]

hi

have a look at Bijoy’s posts

http://www.erlang.com/forum/erlang/thread.htx?thread=4440

and this…

http://www.scribd.com/doc/59598352/17/Speech-Data-Traffic#outer_page_29

[Author]

Posts