Clutter Covariance Matrix Estimation In Heterogeneous Context For STAP

Covariance matrix estimation is a fundamental issue of adaptive signal processing.With the continuous enhancement of radar resolution and the complexity of the clutter environment,performance of clutter sample covariance matrix estimation(SCM)in the traditional Gaussian background will be degraded.Covariance matrix estimation algorithm of heterogeneous clutter can effectively overcome this drawback and perform more robustly with better estimation accuracy.Considering the STAP clutter modeled by compound Gaussian distribution,the estimation of clutter covariance matrix can be viewed as parametric estimation problem of eigenvalue,texture and clutter subspace.In this paper,a two-step maximum likelihood algorithm is proposed to estimate STAP clutter covariance matrix.Firstly,following the maximum likelihood estimation of texture and eigenvalue,eigenvalues' fixed point equation(EFPE)is derived.Secondly,the MajorizationMinimization optimization framework is used to iteratively optimize the subspace of clutter to obtain accurate subspace estimation.Simulation results show that the proposed algorithm can effectively improve the accuracy of clutter covariance matrix estimation,and improve the detection performance of STAP.