Estimation of spectral correlation function based on evolutionary spectral

In this paper, we first demonstrate that a cyclostationary process is an oscillatory process which has a periodic evolutionary spectral, and we determine the relation between spectral correlation function and the periodic evolutionary spectral. Next, based on harmonic series representation (HSR) (a Fourier series with time varying lowpass coefficients representation that reveals the pertinent properties of a cyclostationary process and can cover the vast majority of temporal signals of interest) and discrete prolate spheroidal sequences (DPSS) (the unconditional orthonormal bases which can perfectly represent all lowpass signals such as the time-varying coefficients) and MMSE estimator which is used for estimation of evolutionary periodogram we estimate the spectral correlation function. This approach does not require the data windowing drawback in the time domain and is more robust than conventional cyclostationary approach to varying cyclic frequencies.