Estimation of Direction of Arrival for Two Sinusoidal Frequency Modulated Signals Using Wigner Time-Frequency Distribution in Non-stationary Signal Processing

In non-stationary signal processing, the Wigner time-frequency distribution is frequently employed for signal analysis. This technique can be utilized to estimate the direction of arrival (DOA) for two sinusoidal frequency modulated signals. The Wigner time-frequency distribution serves as a method to characterize signal distribution simultaneously in both time and frequency domains. It decomposes signals into small time-frequency regions, where the average power spectral density within these regions can be used to compute the signal's time-frequency distribution. Implementation typically involves calculating the Wigner distribution using discrete Fourier transforms, with MATLAB code potentially including functions like wvd() for Wigner-Ville distribution computation. The algorithm processes signal segments through quadratic time-frequency representations, where cross-terms between components require special handling for accurate DOA estimation. The enhanced approach may incorporate spatial smoothing techniques and MUSIC algorithm integration to resolve multiple signal sources. Applying Wigner time-frequency distribution in signal processing enables better understanding of signal characteristics and behaviors, particularly for tracking instantaneous frequency variations in non-stationary environments. Code implementation often includes preprocessing steps for signal normalization, time-frequency matrix generation, and peak detection algorithms for DOA extraction from time-frequency representations.