Prev: Fourier transform of an integral
Next: Randy?
From: fisico32 on 30 Apr 2010 11:31
Hello Forum,

the Fourier integral is said to be good only for stationary signals.
Clearly, given a signal f(t) of a certain duration T, we do its FT and find
the composing spectral components.

We could also break the signal into sections and do the FT of each section:
perform the STFT....

IF the FT of each segment is very different, then the signal f(t) is termed
nonstationary...
If the FT of each segment is very similar, the signal is stationary...
How similar do the FTs of different segments need to be actually be to
state stationarity?

Are we comparing only the power spectra of the segments or also the phase
spectra?

thanks,

fisico32



 | 
Pages: 1
Prev: Fourier transform of an integral
Next: Randy?