Classification of radar returns using Wigner-Ville distribution

This paper attempts to analyze and classify the radar returns from a simulated ground based radar using Wigner-Ville distribution (WVD). Stochastic modeling methods like AR and ARMA have earlier been applied for the classification problem. However they suffer from inaccuracies such as showing inexact spectral widths, which is a prime factor to distinguish weather returns from birds, non-adaptability to time-varying clutter spectra, etc., despite being prominent in resolving spectral peaks. The paper proposes WVD as an alternative tool for the classification of radar returns, since it is a powerful technique for time-varying spectra. The analysis shows promising results especially from the view point of spectral widths, though faced with the problem of cross-spectral components. Examples of WVD applied to the simulated radar signal are presented to illustrate the advantages of this method.