Instantaneous frequency estimation using the discrete linear chirp transform and the Wigner distribution

TL;DR

This paper introduces a novel method combining the discrete linear chirp transform and Wigner distribution to accurately estimate instantaneous frequency in noisy, multicomponent signals, reducing cross-term interference.

Contribution

The paper presents a new approach that integrates DLCT and WD for improved instantaneous frequency estimation of non-stationary signals.

Findings

Effective in low SNR conditions

Reduces cross-term interference

Accurately estimates instantaneous frequency

Abstract

In this paper, we propose a new method to estimate instantaneous frequency using a combined approach based on the discrete linear chirp transform (DLCT) and the Wigner distribution (WD). The DLCT locally represents a signal as a superposition of linear chirps while the WD provides maximum energy concentration along the instantaneous frequency in the time-frequency domain for each of the chirps. The developed approach takes advantage of the separation of the linear chirps given by the DLCT, and that for each of them, the WD provides an ideal representation. Combining the WD of the linear chirp components, we obtain a time-frequency representation free of cross-terms that clearly displays the instantaneous frequency. Applying this procedure locally, we obtain an instantaneous frequency estimate of a non-stationary multicomponent signal. The proposed method is illustrated by simulation. The results indicate the method is efficient for the instantaneous frequency estimation of multicomponent signals embedded in noise, even in cases of low signal to noise ratio.