 # Matlab signal detection/analysis

Hello all,

I am attempting to analyze data sets recorded by the USRP in Matlab for
the purpose of signal detection/analysis. The signal detection
requirement must be met by identifying prominent signals and returning
the frequency and signal strength. In doing this I have come to identify
several important questions which I cannot answer myself:
Given a known decimation, how can one determine the sampling rate of an
output data file?
Since the “tuned” frequency is subtracted from the signal, how can one
apply the FFT in Matlab to identify the original signal’s frequency?
What are the units of the data file values?

Jake

On Tue, Jun 24, 2008 at 11:36:34PM +0200, Jake Hertenstein wrote:

apply the FFT in Matlab to identify the original signal’s frequency?
What are the units of the data file values?

Jake

You don’t say, but I’m assuming that the data is being recorded
using usrp_rx_cfile.py. The files contain nothing but the binary
samples. You may want to consider encoding the decimation rate and
center frequency in the filename, or writing the file using something
other than gr.file_sink(…). Note that in general, the guts of GNU
Radio have no clue as the actual sample rate of data streaming through
the blocks. That’s a feature, not a bug.

If the decimation is D, the complex baseband sample rate is 64e6 / D.

Eric

Hello all,

I am a beginner in GNU Radio and I am not an expert in programming, of
sure my question is stupid: can I store signal samples in a file and
import them in MATLAB to implement an energy detector ?
I have not a USRP yet, but firstly I would try to generate a sine curve
and a gaussian noise, sum them and export the samples in MATLAB.
I have seen that exist the file_sink() function but I don’t understand
how it works.

Could you help me ?

Thanks !!!

Riccardo

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.