Research On Non-Stationary Signals Based On Wigner Distribution

Joint time-frequency analysis(JTFA) is a powerful tool in the time-frequency domain to deal with non-stationary signals. The distribution of Wigner-Ville has been payed much attention because of its excellent concentration in the domain of time-frequency and physical properties. As the improved algorithm of WVD transform Radon-WVD transform has not only the excellent characteristics of WVD, but also has the capability of anti noise and cross terms suppression. So it is widely uesd in field of signal processing.This paper discusses the method and performance of parameters estimation of Chirp signal single component and multi-component by Radon-WVD transform in different simulation conditions taking a typical non-stationary signal Chirp as the research object. Besides detecting and analysing the parameters of non-stationary wireless communication signal using WVD transform to verify that the anti-detection performance of non-stationary wireless communication signal. The main work is summerized as follows:For single component of Chirp signal, detection of single component Chirp with Radon-WVD transform can be realized by linear frequency modulation method and image detection method. This paper discusses the effect on perfomance of the parameters estimation in different order of parameters, different sampling rate and time,and the different Gaussian white noise situations. The simulation results show that, the order of parameters does not affect the performance of parameters estimation; on the condition of the under-sampling rate, if the sampling rate is lower, the performance of parameters estimation is worse, but the performance can be improved by increase the length of sampling time; Radon-WVD transforms has a strong anti-noise performance,and its anti- noise ability has athreshold, when the SNR is lower than the threshold,increasing the length of sampling time appropriately can also improve the performance of parameters estimation well.For multi-component Chirp, using time-domain method and frequency-domain method of “CLEAN” idea to realize detection based on Radon-WVD transform.Thispaper sduties different performance and anti-noise ability of the dual Chirp in three kinds of situations of cross, parallel and close to each other. Simulation results show that the frequency domain method of “CLEAN” idea is excellent at detection and has a strong anti-noise performance.For designed non-stationary wireless communication signal, this paper studies the effect that three kinds of modulation including numbers of frequency, period of frequency and the way of frequency changed have on the detection performance based on WVD transform. The simulation results show that WVD transform is able to detect the existance and linear variation characteristics of signal. And WVD transform can get frequency varying sequence through the sliding window detection when the numbers of frequency and the frequency varying rate is low.