Hi,

See this:

https://theses.lib.vt.edu/theses/available/etd-05062011-111048/unrestricted/Malady_AmyC_T_2011.pdf

In the energy detection method you determine the power contained in every frequency component of the detected wireless signal. This can be accomplished by FFT of the discrete time domain of the signal in specific time span represented by the sampling time and the number of samples in the FFT sequence.

In case of cyclo stationary detection  one has to determine the periodicity of the auto correlation function of the received signal.  So one has to determine first the auto correlation of the received signal, then one determines it frequency component to fix its periodicity by using Frourier transform or equivalently FFT. For more information please refer to the publication in the LINK: http://www.ap.ide.titech.ac.jp/publications/Archive/IEICE_TRSR(0707Kimtho).pdf

wish you success

Mainly, there are two algorithms for cyclostationarity detector. The first one is time smoothing domain while the second is frequency smoothing domain. Time smoothing domain is less computationally complex and efficient. In time smoothing algorithms there are two methods, i.e. FFT accumulation algorithm (FAM) and strip spectral correlation algorithm (SSCA).

You can find more details about these algorithms and the MATLAB code in my publications and the other research works which cited has been there. 

May you can visit my profile.

Thank you for your prompt response to the question . I am especially grateful to receive your informations  that really help me for trying to solve the problem . I thank everyone who spent a while  for replying to my question .