Re: [Openbts-discuss] current state of handover

[Dmitri S. <dm...@ma...>]

Hi,

decided to place it to the mailing list, as the following might be interesting
so, 

1) it can be that OpenBTS don't broadcast SI6 for radio-link-timeout information, as the reason for dropping the calls

2) I'll appreciate real-life non-synchronized handover traces, as when only mandatory parameters are included (that seems ok for non-syncd ho), Sagem OT190 performs handover, while Nokia 6150  - don't.
I have nether tried other mobiles, nor did iterations to discover the proper set of parameters/values yet. But such traces could save essential efforts and time 


Regards,
Dmitri





Mon, 3 Sep 2012 09:59:12 +0800 от Jet Wong <icp...@gm...>:
>Hi Dmitri,
>
>
>for the Q1, I do find that call will drop after call been established for some ten or twenty mins for some handsets, I guess this can be that OpenBTS don't broadcast SI6 for radio-link-timeout information. but it is not easy to get a copy of real-life traces of nonsyn-handover .
>
>
>
>
>
>
>as for Q2, I'll double check the sdp.
>
>
>
>
>BRs,
>Jet
> 
>
>
>
On Sun, Sep 2, 2012 at 7:54 AM, Dmitri Soloviev <dm...@ma...> wrote:
>
>>Hi Jet,
>>
>>two problems: 
>>1. different handsets behave in different ways. Seems it is important to find a fail-safe set of optional parameters for Handover Command.
>>I'm lucky that SAGEM OT190 operates properly if only mandatory params are included (as an example, several nokias and Sagem 290 do not jump properly).
>>


The priority is debugging handover logic with this particular handset, and study other phones when the procedure is fixed. 
>>
>>2. please find a SIP trace
>>2.1. when invite message is passed to transaction->MTCInitRTP(), SIP::get_rtp_params() fails when called from SIPEngine::InitRTP()
>>


2.2. ACK looks strange
>>As for the strange ip addresses in FROM and TO - seems it's my fault.
>>
>>So, as for Q1, it would be great if real-life traces of Handover Command for non-synchronized procedure were available
>>


Q.2.1  - if you have any idea, it would be perfect.. To me, SDP is ok.. the only strange thing is the absent of mode (sendrecv), but I dont' think it's a key
>>Q.2.2 - ..
>>
>>I'm sorry for such an incorrect way of asking questions.. Even if I'll git my present state of sources, it would be difficult to repeat the bug due to problem #1.
>>
>>Regards,
>>Dmitri
>>
>>P.S. There were lots of funny things with radio and network parameters when running 2 BTSs ;)
>>
>