All,

Since I’m now involved in the project using the USRP, and am looking at

FFT’s acquired, I recall some mention of a DC component of the resulting

FFT due to ‘hardware’.

What is the general approach that folks have used to deal with this

problem.

Thanks,

John C…

John C. wrote:

All,

Since I’m now involved in the project using the USRP, and am looking at

FFT’s acquired, I recall some mention of a DC component of the resulting

FFT due to ‘hardware’.

What is the general approach that folks have used to deal with this

problem.

Thanks,

John C…

A DC component in the frequency spectrum is from a high average in the

sequence.

If you don’t want to see the DC component try something like the

following

pseudo-code:

#input array - x

#mean is the arithmetic average sum(x)/length(x)

fft(x-mean(x))

cheers,

–Mitchell