Methods of periodically correlated random processes and their generalizations

Abstract

The results obtained by authors in the area of theory and methods of statistical analysis of periodically correlated random processes and their generalizations are presented in this article. The main methods for estimation of their correlation and spectral characteristics: coherent, component, least square method and linear filtration method are analyzed. The ways of generalization of these methods to the case of unknown a priori period of non-stationarity are considered and the possible algorithms of its estimation are presented.