Robust Approach for Clutter Covariance Matrix Estimation with STAP in Heterogeneous Environment

Abstract

The conventional statistical Space-Time Adaptive Processing (STAP) methods, such as sample selection and sample weighting methods, and so forth, have a very low utilization ratio of sample data, which results in that the problem of training samples lack is more prominent in heterogeneous clutter environment. Thus, in this paper, the space-time spectrum of the clutter Cell Under Test (CUT) is estimated according to the distribution characteristics of the clutter and the moving target in the range and space-time two dimensional spectrum plane. In addition, the median filtering is exploited to avoid the disturbance due to the moving target for the estimation of clutter spectrum. Finally, the reconstruction of clutter covariance matrix without sacrificing space-time aperture and clutter suppression is achieved．The results of the simulated experiments demonstrate that the proposed method can effectively alleviate the STAP performance degradation due to the interference target, discrete terrain clutter or isolation interference, compared with the traditional statistical STAP methods.