FFT, Periodogram difference, Power Spectral Density Representation and Effect of Hilbert Trasnform to PSD

37 views (last 30 days)
Hi,
My questions are;
1)Generally,FFT and periodogram do not show same thing! I want to get Power Spectral Density. If signal is real data and in time.
M=1533;fs=27; freqVector = (-(M-1)/2:(M-1)/2)/M*fs;
figure;plot(freqVector,20*log10(abs(fftshift(fft(hilbert(signal))))));
figure;periodogram((hilbert(signal)),[],'onesided',[],27*10^9,'power') why this above line needs abs of it ? why not support complex ?
figure;periodogram((hilbert(back_remvd(1:150,30))),[],'twosided',[],27*10^9,'power') why complex is ok for twosided ?
and the amplitude values btw fft and periodogram representation are not same. Why ?
Can you show me the right PSD and spectrum representation please ?
thanks in advance.

Accepted Answer

Youssef  Khmou
Youssef Khmou on 9 Nov 2013
hans, If you are treating narrow band signals then FFT or pwelch as said by Andrew is sufficient, but for wideband signals, you need to use periodogram.
Periodogram of narrowband signal will show you a narrow line of fundamental frequency along the time axis, for FFT try this function :

More Answers (1)

Andrew
Andrew on 9 Nov 2013
fft and periodogram are not the same things, read help. Try pwelch, it's better for me.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!