- This topic is empty.
13th June 2009 at 06:26 #57131PixGuest
Yes, i got them, but i sent 2 replies… I guess you didn’t receive them ?
I find the results very disappointing 🙂 I just can’t hear any benefit of TFO in HR-HR, which should be where the improvement should be the greatest.
So I think that either there are too many intereferences in the area (I can hear some metallic noise… due to interference ?) or just that TFO was not activated, even though you think it was…
What do you think about your results ?
(Note: non-homogeneous calls CANNOT enjoy TFO… TFO works only if the codec is the same in both MS. And in B10, AMR does not allow TFO, whichever the AMR codec)
Cheers 🙂 and thanks a lot for sharing
Pix13th June 2009 at 18:17 #57132krishGuest
can anybody share ericsson power control and intra cell handover for improving tch drop?14th June 2009 at 07:43 #57133PixGuest
Wasn’t it discussed already ? with t_init, etc. ?15th June 2009 at 11:29 #57134IvannGuest
Sorry for long delay with answer .
There are no letters in ivann-slavinskiy at rambler dot ru 🙁
About TFO testing: yes, I don’t like results too :)Looking through total benchmarking reports everything seemed all right. But analyzing deeper through the logs and obtained results I’ve already found lots of mismatches. I need some time to find out the truth 🙂
About non-homogeneous calls – of course they can’t 🙂
About interference – it’s more than possible because drive-test was made in town with very high traffic and high index of frequency reusing (the next object for optimization for our group 🙂15th June 2009 at 12:15 #57135IvannGuest
sorry I can’t help with Ericsson 🙁
However I think it must be similar to Alcatel and Siemens and other vendors (in according to GSM standarts)15th June 2009 at 14:27 #57136krishGuest
ok ivan and pix can u explain in terms of alacatel?15th June 2009 at 18:20 #57137PixGuest
it’s all explained in this post, check out the past posts.
If you have more questions, read 3GPP about Tandem Free Operations.
Weird about the lost email. Nevermind, I’ll let you do your analysis. On my side, the voice quality samples do not show any difference at all, so I can’t really use them to advertise for TFO activation worldwide.. can I ? 🙂
If you need any alcatel GSM training in Russia, please let me know 🙂
Pix17th June 2009 at 10:04 #57138IvannGuest
about TFO testing – it’s a detective novel 🙂 The most problem is that for 2 BSCs (among 5 in that city) taht very PRC (TFO activation) wasn’t really applied. Therefore now it’s very difficult to seprate correct measurment from incorrect 🙁 There were also found some drawbacks in agilent algorithm of sorting and groupping and futher estimation received samples – now we’re deciding how to compensate it.
We’ll repeat our experiment next week I hope 🙂
about training – thanks a lot, but I’m sorry to say, I’m not from russia – I’m from ukraine.
I don’t know why but we engage training in Moscow very seldom – mainly in romania, Stuttgart and Lannion … But considering economic crisis in our country and company I think we should forget about education abroad for some time :(( In the best case – until B9->B10 migration ….17th June 2009 at 13:07 #57139PixGuest
Ok, so I certainly saw you in Lannion because i work there.
Perhaps, I even trained you… But the only ukranians we remember are the girls :))
For the drive testing, that’s really too bad. Can’t your repeat the measurements from your office ? You don’t really need to drive-test for days, just few minutes in each case should be enough to feel the difference. And you don’t even have to use Agilent… simply record the voice files and trust your ears.
That’s a quick and cheap first step to validate TFO activation, and validate that it could perhaps bring a benefit in VQ… well, i’m just saying… I really want good voice samples :))
At this time of the year, i can see the mountains.
this fashion is attractive to young people.
… i wouldn’t even ask why…18th June 2009 at 09:15 #57140IvannGuest
Of course… Who could remember ordinary ukrainian guys when there were pretty ukrainian girls? I understand you… / just kidding /
I was in Lannion in april-may 2007, we were trained by Mohammed Al-Jafaari (I hope I remember the surname right) – “OMC-R external interfaces.B9”, “NPA operat/admin” and “RNO operat/admin”.So, I think, you can say – could we have met or now 🙂
I hope one day I ‘ll have possibility to thank you personally :))))
…Brittany is wonderful place, it were three unforgettable weeks 🙂18th June 2009 at 16:40 #57141PixGuest
i just resent you an email 🙂 I’m glad you did do a mistake in the surname, we should try to remain anonymous so google would not pickup on his name :)) I’m working with mohamed, you’ll see my name in the email.19th June 2009 at 12:58 #57142IvannGuest
OK, thanks, I’ve received you mail and answered it 🙂30th June 2009 at 20:43 #57143IvannGuest
I’ve just sent you several letters to your mail. They contents results of my last experiment with TFO. I can’t say they arenice or not – you can assess them by yourself 🙂 But now we can be sure that the experiment is performed absolutely correctly, so we can trust these results.
Sorry for very untidy letters.If you have problem with archives let me know and I can repack them and resent.30th June 2009 at 21:55 #57144PixGuest
They are really messy, but I’ll sort it out eventually.
Thanks again for sharing, and for the *absolute perfection* of the methodology of your experiment :))22nd July 2009 at 19:57 #57145IvannGuest
Dear Pix, good evening!
new day brings new problem 🙂
My colleague from other region (Alcatel BSS too) has received complaint from one of our important subscribers. It’s a security firm, they have more than 300 numbers. They use GMS-modems which are located at the objects which are protected by this firm. Once per 1 hour each modem phones to server and transmits a certain tone ringing (I’m not sure I named it correctly – it’s combination of several frequencies (there is a set of such combinations, and each of them encodes a certain messages). It’s a Mesozoic solution :), but it’s our subscriber … they said that month ago they noticed problem: percentage of corrupted ringing increased from ~ 1% to ~30%. Errors appeared and disappeared spontaneously here and there without any visible interconnection. They checked and replaced modems and they also tested used protocols but without any results. After that they came to us.
Considering the fact that beginning of their problem exactly coincides with TFO activation on those BSCs, my colleagues deactivated TFO on 6 BSCs. Subscriber said that problems disappeared.
My question is: Could it be possible (even in theory) that TFO negotiation and TFO commands (which transmits within speech frames) corrupt tone ringings? Or this suggestion is unreal at all ? I don’t like the idea of deactivation TFO for third of BSCs in the region, so the answer to this question is very important to me.
Many many thanks.