Windowed FBSE-EWT method for nonstationary signal analysis

Abstract

Time-frequency (TF) analysis is an active research area in the field of signal processing. It has various methodologies to generate TF representation like short time Fourier transform (STFT), Hilbert-Huang transform (HHT), wavelet transform (WT), Wigner-Ville distribution (WVD), empirical wavelet transform (EWT), etc. which help in getting better knowledge about a signal. In this thesis, we discuss windowed Fourier-Bessel series expansion based empirical wavelet transform (WFBSE- EWT) method for analysis of non-stationary signals which has been developed by enhancing the existing Fourier-Bessel series expansion based empirical wavelet transform (FBSE- EWT) method. It is obtained by segmenting the signal in time domain by using windows such as Gaussian, Hann, Chebyshev and Hamming along with 50% overrun, applying FBSE-EWT on individual segments, adding the resulting intrinsic mode functions (IMFs) and obtaining the TF representation by applying Hilbert transform (HT). This gives us better TF representation as compared to the existing method