|
From: Ralph A. S. d. <ra...@sc...> - 2016-03-18 07:26:43
|
Sure, I can send it later this day, but it is not very special, no hidden tweaks or so :) I changed some paths of some databases, made some small adjustments, but nothing crucial.
Ralph.
From: Simon Pålstam [mailto:sim...@st...]
Sent: Thursday, March 17, 2016 10:25 PM
To: Ralph A. Schmid, dk5ras; ope...@li...
Subject: Re: [Openbts-discuss] Problem with OpenBTS on USRP B210
I don't know if it's a good or bad thing for me that it seems to work for you, but thanks for taking time to write an answer.
I don't think it's necessary to send me the VM, but please copy and send your OpenBTS config if you have time.
Thanks again
/Simon
2016-03-17 16:16 GMT+01:00 Ralph A. Schmid, dk5ras <ra...@sc... <mailto:ra...@sc...> >:
I just can confirm that latest uhd master works with latest OpenBTS quite well, I do not have the issues you describe, although I did not measure the timing error.
My B210 is an older one, with the white PCB, don't know if the green ones behave different.
With two cheap antennae I get without any issues 100-200 m of range, all kinds of phones do connect, Android, iOS, old Nokias, even stuff from 1995 works just fine. Voice quality is what one can expect from FR codec, calls get connected quite reliable, call setup does not fail very often.
Only GPRS is unstable, but I blame my virtual machine for this.
No GPSDO here, just the standard oscillator.
All in all it is a reliable setup.
I could share the VMware VM, there is no personal stuff in it, but I'm afraid it is a quite large one, some 30 or 40 GB, and I only can offer 10 Mbps of uplink speed.
Ralph.
From: Simon Pålstam [mailto:sim...@st... <mailto:sim...@st...> ]
Sent: Thursday, March 17, 2016 3:32 PM
To: ope...@li... <mailto:ope...@li...>
Subject: [Openbts-discuss] Problem with OpenBTS on USRP B210
Hi, I'm trying to set up OpenBTS on an USRP B210 with installed GPSDO and have run into the following problems:
1. It is impossible to connect to the network with Android phones
2. The quality of service when making a call with an Iphone is really bad
3. The range at which it is possible to make a call is maximum 30 meters, even with maximum power
4. The measured timing error on the channel is always at least -1 bit period
5. The measured timing error seem to be drifting, resulting in a saw tooth like graph. The deviations between the top and the bottom of the saw tooth pattern usually about 0.4 bit periods.
6. The system seem generally unstable and many results are inconsistent
So far I have installed a GPSDO to get a better clock, but without any improvement on the OpenBTS network. The GPS lock indicator is illuminated and everything seems fine when I query the GPSDO.
It would be really nice to get an expert opinion on what might be the problem. And if some of these phenomenons are to be expected and simply reflects the performance of the system.
Sorry if it seems like I'm spamming the mail-list but I have already spent way to much time on getting this to work
I have heard that the problem could be that the current version of UHD is ahead of OpenBTS in the development and thus some changes has not been compensated for in OpenBTS. If this is the case, do anyone know a combination of UHD and OpenBTS versions that is confirmed to work together?
Thanks for any help
/Simon
|