Research On Chirp Parameter Estimation Based On RPCA

Robust principal component analysis(RPCA)is an important component in the framework of low rank matrix restoration theory.It mainly solves the problem of low rank part and sparse part decomposition in observation matrix.This algorithm is widely used in many field.Chirp signal is one of the most important time-varying signals in the field of signal processing.It is widely used in radar,sonar,communication,biology and geological prospecting.Therefore,Chirp signal parameters estimation is an inportant task for single channel passive detection system.The most popular method for estimating parameters of Chirp signal is time-frequency analysis based method.This method can describe the distribution characteristics of signal in time-frequency domain,which makes system easy to estimate parameters of non-stationary signal.Although there are many methods can estimate Chirp signal parameters,most of them do not solve the problem of parameter estimation when interference exist.In this paper,we study how to estimate Chirp parameters for single channel passive detection system when Chirp signal in the data overlap with Phase-Shift Key(PSK)interference in time and frequency domains.The main work and innovation of this paper are as follows:Firstly,the mathematical model of RPCA is studied.This paper introduce several algorithms can solving RPCA difficulty,including accelerated approximation gradient(APG)algorithm,Augmented Lagrangian multiplier(ALM)method and Singular Value Threshold(SVT)method.The implementation steps of the above algorithm are given,and the algorithm is simulated and analyzed.Then,the mathematic model and basic properties of Chirp signal are introduced.A variety of Chirp signal parameter estimation methods are also studied,including STFT and least squares based method,WVD and Hough transform based method,Fr FT and peak detection based method.The implementation steps of the above algorithm are given,and the algorithm is simulated and analyzed.Finally,we try to solve the Chirp signal parameter estimation problem when Chirp signal and PSK signal overlapped in time and frequency domains.This paper presents a parameter estimation method based on RPCA.The technique involves,firstly,the received signal is transformed onto a time-frequency(TF)distribution matrix by the short-time Fourier transform(STFT).Then using Hankel structured matrices constructed by frequency vectors of the TF matrix,the PSK interference excision is solved by RPCA.Finally,the relative parameters of chirp signal are estimated using the processed well-focused TF matrix by linear fitting method.Experimental results of simulated and real data have demonstrated the effectiveness of the proposed method.