LFM signal analysis based on synchrosqueezing-Hough transform

Abstract

The synchrosqueezing transform is a time-frequency (TF) analysis tool to process non-stationary signals. Unfortunately, it does not produce accurate TF results when faced with LFM signals. In this paper, we propose synchrosqueezing-Hough transform (SS-HT) to address this problem. First, we introduce Hough transform to obtain more accurate instantaneous frequency (IF) estimation. Then, in order to improve the TF concentration, a new TF rearrangement operator is constructed. Furthermore, we demonstrated that SS-HT has the ability to reconstruct the signal. The experimental results prove that SSHT can not only obtain high TF concentration, but also improve the accuracy of IF estimation and parameters estimation of LFM signals.