gr-gsm vs. airprobe with bladeRF

Having issues with the site, hardware, source code, or any other issues?
Post Reply
blrfuser
Posts: 9
Joined: Tue Oct 13, 2015 11:21 am

gr-gsm vs. airprobe with bladeRF

Post by blrfuser »

Hello,

I successfully followed these instructions:

https://gist.github.com/sovietw0rm/7540c86f66e447480ee7

to use airprobe with bladerf - the only modification to those instructions was to use a different git address for 'airprobe' since the gnumonks git repo no longer exists...

Anyway, it works. I have airprobe running with my bladeRF. However, I think it is very old - the interface does not look at all like the gr-gsm tool that I see used in documentation and tutorials/examples and there is very little control over the tuning and signal.

So I think I need to follow a similar set of instructions, but with gr-gsm instead of airprobe ... but I cannot find such instructions.

Is this patch:

https://github.com/scateu/airprobe-3.7-hackrf-patch

still applicable to gr-gsm, or perhaps it is already built in ?

Does anyone know of a howto/instructions to get gr-gsm working with the bladerf ?

Thank you.
blrfuser
Posts: 9
Joined: Tue Oct 13, 2015 11:21 am

Re: gr-gsm vs. airprobe with bladeRF

Post by blrfuser »

Sorry. Ignore. gr-gsm right out of the box appears to work just fine with bladeRF (at least with a modern gnuradio environment in 2015).

Not sure why I thought it wasn't working - there was no reason at all for me to patch airprobe or any of that.

Thanks.
blrfuser
Posts: 9
Joined: Tue Oct 13, 2015 11:21 am

Re: gr-gsm vs. airprobe with bladeRF

Post by blrfuser »

replying to my own thread here just to give future readers additional information ...

First, kalibrate-bladerf really does not work at all. It appears to "find" all kinds of GSM towers, but really is just generating false positives. I manually scanned all of 850GSM and PCS bands and found one cingular tower and two TMobile towers (which makes sense) but kalibrate-bladerf had a long list of 7-8 "towers" ... none of them existed in real life *and* kalibrate-bladerf did not find any of the real towers.

So I would not recommend relying on kalibrate-bladerf for anything and I suspect it is not usable until rewritten or debugged.

Second, airprobe/gr-gsm, when used with bladerf, cannot tune below 925mhz. This is because there is actually a hard-coded frequency floor of 925mhz around line number 86 in the python script ... you can just edit that and change it to 800, which I did. That allowed me to tune GSM base stations in the 850GSM band.

Hope that helps.
jynik
Posts: 455
Joined: Thu Jun 06, 2013 8:15 pm

Re: gr-gsm vs. airprobe with bladeRF

Post by jynik »

Thanks for sharing your findings!
blrfuser wrote:So I would not recommend relying on kalibrate-bladerf for anything and I suspect it is not usable until rewritten or debugged.
I second that. I've posted one required change here on its issue tracker. There are probably many more, but I haven't reviewed or looked at this code much.
blrfuser wrote:Second, airprobe/gr-gsm, when used with bladerf, cannot tune below 925mhz. This is because there is actually a hard-coded frequency floor of 925mhz around line number 86 in the python script ... you can just edit that and change it to 800, which I did. That allowed me to tune GSM base stations in the 850GSM band.
Any chance of getting a pull request submitted for this?
Post Reply