A new method for non-stationary signal analysis using eigenvalue decomposition of the Hankel matrix and Hilbert transform

Abstract

Non-stationary signal analysis is an essential part for many engineering fields. Time-frequency analysis methods are commonly used methods for analysis of non-stationary signals. In this paper, a new domain for time-frequency analysis has been proposed which has been studied for the analysis of non-stationary signals. The proposed method combines two techniques namely, iterative eigenvalue decomposition of the Hankel matrix (IEVD-HM) and the Hilbert transform (HT). The IEVD-HM technique provides a set of mono-component signals where the HT has been employed to determine amplitude envelopes and instantaneous frequencies of these mono-component signals. These amplitude envelope and instantaneous frequency estimations have been used to determine the time-frequency representation. The obtained time-frequency representation has been studied for the analysis of synthetic non-stationary signals in order to show the effectiveness of the proposed method. © 2017 IEEE.