Writing RSSI Values to a File Twice/Sec

Hi all,

I’m new to GNURadio and I’m trying to use a USRP E310 to capture GSM
signals from a cellphone and store the RSSI values in a text file. If
you take a look at my flow graph I’ve managed to capture the average
strength across the 200kHz GSM band and I’m trying to filter out the
unwanted signals using the Probe Avg Mag^2 block.

I’m having trouble writing to the file sink at a constant 2 samp/sec.
Could you please give me advice on how I should fix my flow graph?

Thank you,

Varun

On 07/02/2015 11:07 AM, Varun Nambiar wrote:

Hi all,

I’m new to GNURadio and I’m trying to use a USRP E310 to capture GSM signals
from a cellphone and store the RSSI values in a text file. If you take a look at
my flow graph I’ve managed to capture the average strength across the 200kHz GSM
band and I’m trying to filter out the unwanted signals using the Probe Avg Mag^2
block.

I’m having trouble writing to the file sink at a constant 2 samp/sec. Could you
please give me advice on how I should fix my flow graph?

What kind of trouble? Can you also post the grc file so I can try it
locally.

Also, which file system image are you using? The one from the card that
came with the unit, or release-2?

Philip

On 07/02/2015 11:25 AM, Varun Nambiar wrote:

Hi Philip,

At the moment of execution a file size of 32KB with just 0s is written to my
file and then it doesn’t add anymore values after that. I am on release-2 and I’ve
attached my GRC file.

I dropped the throttle block and saw the output file get larger. Resend
the flowgraph to the list and maybe someone else has some ideas how to
make it better.

Philip

Hi all,

Here’s the GRC file. Any help would be appreciated!

Thank you,
Varun


From: Philip B. [email protected]
Sent: Thursday, July 2, 2015 1:35 PM
To: Varun Nambiar; GNURadio D.ion List
Subject: Re: [Discuss-gnuradio] Writing RSSI Values to a File Twice/Sec

On 07/02/2015 11:25 AM, Varun Nambiar wrote:

Hi Philip,

At the moment of execution a file size of 32KB with just 0s is written to my
file and then it doesn’t add anymore values after that. I am on release-2 and I’ve
attached my GRC file.

I dropped the throttle block and saw the output file get larger. Resend
the flowgraph to the list and maybe someone else has some ideas how to
make it better.

Philip

On 07/02/2015 03:26 PM, Varun Nambiar wrote:

Hi all,

Here’s the GRC file. Any help would be appreciated!

Thank you,
Varun
Could you clarify whether you want RSSI estimates over the whole input
bandwidth, or for each discrete frequency bin?

This graph shows confusion in a number of areas, so, what is it you’re
trying to do?

Hey Marcus,

I’m getting the FFT over a 1 MHz bandwidth and I’m computing the average
signal strength over a 125 kHz region of that 1 MHz region by
integrating all the samples and dividing by the number of samples. I’m
hoping to get an RSSI estimate over the input bandwidth. My issue is
that I can’t write to the file at 2 samples/sec. When I execute the code
in the attached GRC, at the moment of execution I get a 32 KB file with
0s and it doesn’t output any samples to the file. When the throttle
block is removed, the processor on the E310 writes to the file at the
clock speed which leads to incredibly large files. Is there a way to cut
it down to 2 samples/sec? Also the Avg Mag^2 block is there to probe for
signals of strength greater than -70 dB.

Thanks,
Varun


From: discuss-gnuradio-bounces+vnambiar=removed_email_address@domain.invalid
discuss-gnuradio-bounces+vnambiar=removed_email_address@domain.invalid on behalf of Marcus
D. Leech [email protected]
Sent: Thursday, July 2, 2015 3:56 PM
To: [email protected]
Subject: Re: [Discuss-gnuradio] Writing RSSI Values to a File Twice/Sec

On 07/02/2015 03:26 PM, Varun Nambiar wrote:

Hi all,

Here’s the GRC file. Any help would be appreciated!

Thank you,
Varun

Could you clarify whether you want RSSI estimates over the whole input
bandwidth, or for each discrete frequency bin?

This graph shows confusion in a number of areas, so, what is it you’re
trying to do?


From: Philip B. [email protected]mailto:[email protected]
Sent: Thursday, July 2, 2015 1:35 PM
To: Varun Nambiar; GNURadio D.ion List
Subject: Re: [Discuss-gnuradio] Writing RSSI Values to a File Twice/Sec

On 07/02/2015 11:25 AM, Varun Nambiar wrote:

Hi Philip,

At the moment of execution a file size of 32KB with just 0s is written
to my file and then it doesn’t add anymore values after that. I am on
release-2 and I’ve attached my GRC file.

I dropped the throttle block and saw the output file get larger. Resend
the flowgraph to the list and maybe someone else has some ideas how to
make it better.

Philip

Thank you,
Varun


From: Philip B. [email protected]mailto:[email protected]
Sent: Thursday, July 2, 2015 11:18 AM
To: Varun Nambiar;
[email protected]mailto:[email protected]
Subject: Re: [Discuss-gnuradio] Writing RSSI Values to a File Twice/Sec

On 07/02/2015 11:07 AM, Varun Nambiar wrote:

Hi all,

I’m new to GNURadio and I’m trying to use a USRP E310 to capture GSM
signals from a cellphone and store the RSSI values in a text file. If
you take a look at my flow graph I’ve managed to capture the average
strength across the 200kHz GSM band and I’m trying to filter out the
unwanted signals using the Probe Avg Mag^2 block.

I’m having trouble writing to the file sink at a constant 2 samp/sec.
Could you please give me advice on how I should fix my flow graph?

What kind of trouble? Can you also post the grc file so I can try it
locally.

Also, which file system image are you using? The one from the card that
came with the unit, or release-2?

Philip

Thank you,

Varun


Discuss-gnuradio mailing list
[email protected]mailto:[email protected]
https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.gnu.org_mailman_listinfo_discuss-2Dgnuradio&d=AwIC-g&c=1QsCMERiq7JOmEnKpsSyjg&r=oML8w7v4Y02f11gsS3bWKQ&m=hC1yMZO9YJC94wCgIZWYH9bbke_ocPJKoWRVfAMsuMo&s=8BPBpMr_86l6B-6kzl20V_C-fjCtQ-eOYPqCNrrvwCE&e=

On 07/02/2015 05:51 PM, Varun Nambiar wrote:

file with 0s and it doesn’t output any samples to the file. When the
throttle block is removed, the processor on the E310 writes to the
file at the clock speed which leads to incredibly large files. Is
there a way to cut it down to 2 samples/sec? Also the Avg Mag^2 block
is there to probe for signals of strength greater than -70 dB.

Thanks,
Varun

Why not:

signal-source—>filter---->complex-to-mag^2---->single-pole-iir-filter—>keep-one-in-N----->file-sink

Adjust your post-detection filtering and keep-one-in-N to be consistent
with your desired logging rate.