Novel Fourier Quadrature Transforms and Analytic Signal Representations for Nonlinear and Non-stationary Time Series Analysis

TL;DR

This paper introduces novel Fourier quadrature transforms and analytic signal representations that enhance the analysis of nonlinear and non-stationary signals, offering improved time-frequency-energy insights.

Contribution

It proposes new Fourier quadrature transforms and analytic signals, extending the Fourier decomposition method for better nonlinear and non-stationary signal analysis.

Findings

FQTs produce Fourier spectrum with only positive frequencies.

FQAS representations have real or imaginary parts as the original signal.

Enhanced time-frequency-energy analysis of real-life signals.

Abstract

The Hilbert transform (HT) and associated Gabor analytic signal (GAS) representation are well-known and widely used mathematical formulations for modeling and analysis of signals in various applications. In this study, like the HT, to obtain the quadrature component of a signal, we propose novel discrete Fourier cosine quadrature transforms (FCQTs) and discrete Fourier sine quadrature transforms (FSQTs), designated as Fourier quadrature transforms (FQTs). Using these FQTs, we propose sixteen Fourier quadrature analytic signal (FQAS) representations with following properties: (1) real part of eight FQAS representations is the original signal and imaginary part of each representation is FCQT of real part, (2) imaginary part of eight FQAS representations is the original signal and real part of each representation is FSQT of imaginary part, (3) like the GAS, Fourier spectrum of the all FQAS representations has only positive frequencies, however unlike the GAS, real and imaginary parts of FQAS representations are not orthogonal. The Fourier decomposition method (FDM) is an adaptive data analysis approach to decompose a signal into a set of Fourier intrinsic band functions. This study also proposes new formulations of the FDM using discrete cosine transform with GAS and FQAS representations, and demonstrates its efficacy for improved time-frequency-energy representation and analysis of many real-life nonlinear and non-stationary signals.