Elimination of cross‐terms in the Wigner–Ville distribution of multi‐component LFM signals

Abstract

A novel method is put forward to remove cross‐terms in the Wigner–Ville distribution (WVD) of multicomponent linear frequency modulation (LFM) signals. For the difference of the amplitude of auto‐terms and cross‐terms, the WVD matrix is filtered using low‐pass filter in the direction of frequency modulation ratio. Furthermore, a novel technique called energy weight is proposed for the cross‐terms with low‐oscillation frequency which cannot be eliminated because of the limitation of filters. The gain of signal to noise ratio (SNR) after filtering is derived while the resolution and computational cost are also analysed. The simulation and experimental results prove that this method improves the SNR along with the removal of cross‐terms with maintaining high resolution.