Abstract
Wavelet autocorrelation function is proposed in this paper. After Cooley and Tukey developed the algorithm of fast fourier transform, the power spectral density is used in various field to obtain the frequency components of stationary signals. Autocorrelation function is derived as inverse fourier transform of the power spectral density. In the autocorrelation function, the noise components of target signal are reduced, although the periodic components still remain. In this study, autocorrelation function of wavelet transform is proposed for time series analysis of nonstationary signal, and its ability for noise suppression is investigated.
