The output of spectrum_sense must be shifted to minus

GNU Radio
1

Im trying to sense the noise variance when there is no signal.
I’m consider a band from 1.2GHz to 1.25GHz.
Ofcourse the output is not the same for all frequencies sweeping parts.
However when there is no signal i get reading flactuates between 10dB to
20
dB, which, according to the spectrum analyzer, should be around -90dB.
While when there is a signal i get readings fluctuates around 80 dB,
which
in reality much lower than this (around 0dBm)

Therefore i think there is a shift should be done for the output of the
FFT.
Any body know how much?

regards,

–
View this message in context:
http://old.nabble.com/The-output-of-spectrum_sense-must-be-shifted-to-minus-tp27026849p27026849.html
Sent from the GnuRadio mailing list archive at Nabble.com.

2

On Wed, Jan 6, 2010 at 05:06, abbasi9999 [email protected] wrote:

Therefore i think there is a shift should be done for the output of the FFT.
Any body know how much?

In general, this will need to be calibrated, and is dependent on
individual USRP variations, the daughterboard gain, and decimation
settings used with the USRP. It is possible when using usrp_fft.py to
use the --ref-scale option to set the equivalent dBFS sample
amplitude, which will then set 0 dB in the display to that amplitude.

Johnathan