A novel method to detect almost cyclostationary structure

Abstract

This paper is devoted to establish a computational approach to investigate that a discrete-time almost cyclostationary model is a suitable choice to fit on an observed dataset. The main idea is estimating the support of spectra and applying multiple testing. The simulated and real datasets are applied to study the performance of the introduced approach. The results confirm that the presented method acts efficiently in view of power study.

Keywords

• Spectral analysis
• Detection
• Almost periodically correlated
• Almost cyclostationary
• Discrete Fourier transform
• Multiple testing