I am currently working on pulsar signal periodicity detection. Most of the literature which i studied used traditional FFT (Fast Fourier Transform) or (Epoch Folding) techniques to find the periodicity of pulsar signal. I am here to ask that do we have any other algorithms which we can use to find the periodicity of signal in frequency domain or in time domain ?looking for your kind help and suggestions.
3
-
$\begingroup$ Welcome to SE.SP! Please read about the AMDF. It might help. $\endgroup$– Peter K. ♦Sep 12, 2021 at 19:11
-
$\begingroup$ Those of us in the audio and music signal processing field call this problem "pitch detection". $\endgroup$– robert bristow-johnsonSep 13, 2021 at 1:34
-
$\begingroup$ I think you are looking for spectral estimation. Traditional methods use FFT and these methods suffer the frequency resolution limitation of FFT, while modern spectral estimation such as MUSIC is kinda "super-resolution" method which is more popular in the recent decades. $\endgroup$– ZR HanSep 13, 2021 at 1:40
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
In the time domain, one can use auto-correlation function to find the period of a signal given the signal duration is longer that the period of the signal. To avoid uncertainty due to damped oscillation, one can clip the signal before computing the auto-correlation.
-
$\begingroup$ //... one can clip the signal before computing the auto-correlation. // oooh, that's not such a good idea. $\endgroup$ Sep 13, 2021 at 1:32
-
$\begingroup$ What i want to do is look for any signal which is periodic or not in the frequency domain or in the time domain, i have tried autocorrelation and FFT it workss well and I get the periodicity of the signal but these are the traditional way to find the periodicity of a signal. $\endgroup$– TariqSep 13, 2021 at 4:21