[USRP-users] Problems using Airprobe with USRP1

GSM Research gsm.research.group at gmail.com
Thu Apr 11 15:36:54 EDT 2013


On Thu, Apr 11, 2013 at 2:51 PM, Ian Buckley <ianb at ionconcepts.com> wrote:

>
> hexdumping a file full of binary complex floats is going to make far more
> sense with some command line options:
>
> hexdump -e '1/4 "%f" "  "1/4 "%f" "\n"' output.cfile
>

Thanks Ian, that does make the data in the files look more similar


> Explicitly setting the sample rate used to capture the off-air data is
> probably a tremendously good idea, I believe for USRP1 "-d 112" is what is
> typically used for Airprobe but you should check that.
>

I agree. But although the usrp_rx_cfile.py program had the -d option, the
same option doesn't exist on the uhd_rx_cfile program.  It *does* offer a
samp-rate option, but we are puzzled on how to set that value to get the
proper decimation.  We have tried values around 571428 (which is close to
64MHz / 112), but it doesn't seem to have any effect on the output we are
getting for gsm_receive.py


> Starting with the RX gain low and working up would be a good methodology,
> starting with it turned way up is a guarantee that the RFX900 is operating
> in a non-linear range.
>

We have been trying different values for the gain.  The run I pasted into
the email is just one that was run recently. But, I guess we should stick
to the lower values for the time being.

Also:

On Thu, Apr 11, 2013 at 12:46 PM, Marcus D. Leech <mleech at ripnet.com> wrote:
   > I'd suggest contacting the folks who wrote gsm_receive.py, who are
unlikely to be adequately represented here.

Well unfortunately, there doesn't seem to be much activity for Airprobe
anymore, and the mailing list doesn't appear to allow new subscriptions
(not that there would be much activity anyway).  The USRP-users list
appears to have the most recent discussions regarding Airprobe, so this is
where I have come in my quest for information.  However, perhaps I will
also contact the folks in the AUTHORS file.

   > uhd_rx_cfile produces a file containing complex-floats representing
the I and Q components of the sampled signal.

That is good to know and we didn't know that.  Thanks Marcus!

John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ettus.com/pipermail/usrp-users_lists.ettus.com/attachments/20130411/b2008243/attachment-0002.html>


More information about the USRP-users mailing list