The Clutter Suppression And Target Detection Based On Polarization Space Time Processing

In recent decades,using polarization information to improve the ability of radar de-tection,anti-jamming and recognition is an important topic in radar,which has been widely concerned by scholars.With the introduction of polarization,the signal dimension also increases.In order to characterize the clutter statistical feature including the polarization feature,more training samples are needed.In practice,the radar needs to face a com-plex and changeable environment for the target detection,which leads to the heteroge-neous clutter and the number of suitable clutter samples in the adjacent range rings are often small,so it is particularly important to improve the detection performance of the polarization radar in the complex clutter environment.This thesis focuses on the clutter suppression and target detection.Based on the polarization space-time adaptive process-ing technology,the problems of polarization clutter covariance matrix(CM)estimation,transmitted polarization design and reduced-dimension processing are studied.The main work and contributions of this thesis are as follows(1)This thesis discusses the general model of signal and clutter in the polarization-space-time domain based on the feature of polarization,and proposes a direction-of-arrival(DOA)estimation algorithm based on the polarization array.Then,the thesis introduces the polarization radar signal processing about the clutter suppression and target detection,and gives a cascade structure.(2)This thesis investigates the robust estimation of Kronecker-structured covariance matrices for compound Gaussian data.Firstly,the compound Gaussian samples are trans-formed into the corresponding complex angular Gaussian samples,and the method of adding a penalty term based on Kullback-Leibler divergence to their negative log-likelihood function is proposed.A covariance matrix estimator based on the fixed-point equation is derived,which is called the robust shrinkage Kronecker estimator(RSKE).Then,the suf-ficient conditions for the existence of the estimator are discussed.An iterative algorithm for solving the corresponding fixed-point equation is proposed.The convergence of the iterative algorithm is proved.This thesis finally introduces three methods to obtain the shrinkage factors by exploiting oracle approximating shrinkage(OAS)and cross valida-tion(CV).Simulation results show that the proposed RSKE has better performance in the estimation accuracy than the existing methods.(3)In order to enhance the interference(clutter)suppression performance of MVDR beamformer,this paper studies the shrinkage coefficient selection of shrinkage sample covariance matrix(S~2CM)and shrinkage Tyler's estimator(STE)for Gaussian and com-pound Gaussian data,respectively.In this thesis,starting from the Gaussian distribution,we utilize the output power of the MVDR beamformer as the objective function,and de-rive a shrinkage coefficient selection algorithm based on the cross validation(S~2CM-CV),which has a low computational complexity.Then,we present an improved shrinkage co-efficient selection algorithm(S~2CM-AE)to further reduce the computational complexity,which is asymptotically approximated to S~2CM-CV.Finally,these algorithms are extended to the compound Gaussian case,and STE-CV and STE-AE algorithms are obtained.Com-pared with some existing algorithms,the proposed algorithm has more extensive applica-bility or better performance.(4)This thesis considers the performance optimization of space-time adaptive pro-cessing(STAP)based on a diversely polarized antenna(DPA)array.A joint transmitted polarization and received filter design to enhance the detection performance is proposed.Accounting for the special structure of the polarization signal,this thesis formulates the problem as a quasi-convex form based on the block majorization-minimization(MM)prin-ciple and solve it in an iterative framework.Compared with several existing methods,the proposed method can achieve better clutter suppression performance.This thesis analyzes the performance of conventional STAP and polarization-STAP(PSTAP),which indicates that the polarization diversity can help improve the performance of STAP for targets with low velocities,especially stationary ones.(5)This thesis considers the problem of the dimensional reduction in the angle-Doppler domain.A recursive algorithm for selecting the best angle-Doppler channels that maxi-mize the output signal-to-clutter-plus-noise ratio(SCNR)is proposed.In each step of the algorithm,the channel which maximize the output SCNR is selected from the remain-ing channels.Compared with several existing STAP methods,the proposed method can achieve better clutter suppression performance at lower computational complexities when the degrees of freedom(DoFs)of the STAP system are fixed.Moreover,the proposed method can help address the issue of training data shortage,which may be particularly attractive for heterogeneous scenarios.