[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Discuss-gnuradio] measure of the SNR in a packet

From: Matteo Campanella
Subject: Re: [Discuss-gnuradio] measure of the SNR in a packet
Date: Wed, 26 Apr 2006 21:50:28 +0200

The first method looks interesting, the only problem I find with it is that I cannot tell wether what I am getting from the channel is noise or noise plus some weak signal I cannot identify as such. When it comes to very busy channels (eg on APRS frequency, you have signals almost all of the time...)

----- Original Message ----- From: "Greg Troxel" <address@hidden>
To: "Matteo Campanella" <address@hidden>
Cc: <address@hidden>
Sent: Wednesday, April 26, 2006 2:47 PM
Subject: Re: [Discuss-gnuradio] measure of the SNR in a packet

 Hello, I would like to write some code to calculate the SNR of a signal;
 in my case this signal is a packet radio AX25 1200 baud packet burst
 received thru TV RX and usrp.

First you need to define SNR.  If you measure the noise power of the
background within the passband when there is no signal, and then
measure the power of the signal+noise while receiving, you can
calculate an estimate.

Or, you can do coherent demod of the signal, and then regenerate a
perfect signal at an amplitude that best matches the received data.
Then you can subtract that and look at the residuals, and calculate
power of the reconstructed signal and the residuals.  The residuals
will have not only noise and receiver distortion, but also components
due to amplitude variations over the estimation period.

This is probably much easier if you don't try to get accurate values
at low SNR.  But it's down low that it's a more interesting problem,
at least to me.

       Greg Troxel <address@hidden>

reply via email to

[Prev in Thread] Current Thread [Next in Thread]