- This topic has 67 replies, 1 voice, and was last updated 11 years, 4 months ago by Peace.
27th May 2009 at 11:29 #57116IvannGuest
at last we have results of our experiment – they are excellent! For this BSC percentage of MC561 amongst other TCH assignment failures (MC541a, MC551, MC561 and MC812) is less than 6% – it’s very good value!! For example, right after TFO activation this value was ~ 26%. We get good KPI even for nonhomogenus cells. Once again thanks a million !!
🙂 Now the most interesting thing for me is what benefit in MOS shall we get after activation TFO for the whole territory 🙂 We’re making drive-tests right now (without TFO) and then we’ll do it again after activation 🙂27th May 2009 at 12:11 #57117PixGuest
I’m very happy for you 🙂 I miss working on the field 🙂
So basically, you should see 2 scenarios:
1/ In loaded cells, calls are setup in HR. If the call is MS to MS, then the distant MS will be downgraded to HR (if distant cell allows it) and TFO will be ON.
2/ In non loaded cells, MS-MS calls, both MS should be in EFR.
After HO, if one MS goes to HR, the distant MS shall be downgraded to HR.
During your drivetests, you should ensure that HR-HR calls with TFO have better voice quality than without (this, you must check by listening to the mobile phone DURING the drivetest… SQI is not so reliable)
Don’t forget that MS-Fixed line will not use TFO 🙂29th May 2009 at 20:24 #57118IvannGuest
Dear Pix, results are excellent! Our KPIs are within limits.
I’m going to set these parameters for the whole network.
Things are worse with drive-tests 🙁 – we’ve finished the first part (without TFO) and got MOS = 3,1 (not bad but also not very good). Of course, we were measuring in MS-to-MS mode 🙂 As for measuring voice quality I’m absolutely satisfied with our new tool – Agilent platform. It’s very flexible and reliable tool. And its huge advantage is that it stores recived audio samples – so it’s possible to listen to them after drive test to assess voice quality by yourself.
I hope that within a few days we’ll have the other part of results – MOS with TFO :)) I’m looking forward to it.
P.S. Yes, working on the field is really very interesting and creative :))30th May 2009 at 08:01 #57119PixGuest
If I can ask you something : could you send me an audio sample of HR without TFO and another of HR with TFO ?
I would love to hear the difference.
Let me know if that’s possible 😉30th May 2009 at 23:36 #57120RexGuest
Hi Pix, Ivann.
What about the case:
FORCE_TFO_HR_WHEN_LOADED=HR not forced
FORCE_TFO_VS_AMR = DISABLE
KEEP_CODEC_HO = free choice of the codec
Would TFO work in this case? And how?
Rex31st May 2009 at 08:27 #57121PixGuest
It would work, but most MS in loaded cells that should establish the call in HR are going to be established in FR.
This is because the HR is not forced, so probably, it’s not going to be chosen, ever.31st May 2009 at 21:38 #57122RexGuest
we owe you so much regarding your contribution in this forum. We just can thank you, nothing else. ALU should know that and give a reward to you and to this forum. I realy mean that.
Regards,1st June 2009 at 08:49 #57123WithneyGuest
Hi all !!
Sorry bu where exactely are these parameters set?1st June 2009 at 10:13 #57124PixGuest
Rex, thanks 😉
Whitney, in the 1353 RA OMC-R.1st June 2009 at 21:56 #57125IvannGuest
sorry for a long pause – I have a business trip right now.
But for the first – I’m absolutely agreed with Rex 🙂 Your contribution in this forum and your help is really inappreciable!
For the second – of course, no problem with samples. But I can do it only next week 🙁 And some more detailes – are you interested in samples without TFO with different codecs on two sides (FR – HR for example) and samples with TFO when HR is used on both sides?
I’ll sort all samples with/without TFO according to used combination ( EFR – EFR, HR – HR, EFR – HR, e.g) – so there will be a huge amount of material.2nd June 2009 at 05:33 #57126PixGuest
Well, you know me (you can guess), I’m interested in all that stuff, definitely.
Thanks, and enjoy your trip,
Pix2nd June 2009 at 08:47 #57127asoGuest
I have problem that using Huawei BSC and aclactel MSC when AMR activated on huawei BSC and when later check the AMR is not work then made trace found that AMR command is send to huawei BSC but from Huawei BSC to MSC the AMR command is not accepted can you hep.10th June 2009 at 12:55 #57128IvannGuest
At last I’m in office and I’ve almost sorted voice samples from our drive-tests (with/without TFO) in according to mode (HR-HR, HR-FR, FR-FR etc.)
Total volume of archive is less than 15 MB. I can send it to you.10th June 2009 at 15:17 #57129PixGuest
i don’t know if you can fit all in one mail, perhaps better to split 5Mb / mail.
pix_erlang at yahoo dot fr
thanks a lot !
pix12th June 2009 at 20:00 #57130IvannGuest
I sent two letters yesterday – I hope you’ve received them both (and all archives are OK I hope).
If no – I can easily repeate it(them).