Exploiting the Persymmetric Property of Covariance Matrices for Knowledge-Aided Space-Time Adaptive Processing

Abstract

In space–time adaptive processing (STAP) technique, the estimation of the interference-plus-noise covariance matrix is one of the critical points. Incorporating a priori knowledge into STAP architectures can reduce the effect of the heterogeneous environment and substantially improve the estimation accuracy of the covariance matrix. Besides the prior information, the persymmetric structure in radar systems with symmetric spaced linear array and constant pulse repetition interval can also be exploited to improve the STAP performance. In this paper, we present a new computationally adaptive knowledge-aided STAP method that requires fewer samples by utilizing the persymmetric structure of the covariance matrix. In addition, based on the covariance matrix estimation technology of the newly proposed knowledge-aided STAP method, two knowledge-aided persymmetric adaptive detectors in the nonhomogeneous environment are proposed as well. First, a two-step design procedure-based detector is proposed for the partially homogeneous model, which is called knowledge-aided persymmetric adaptive coherence estimator. Second, we improve the stochastic heterogeneous model and propose a new knowledge-aided persymmetric generalized likelihood ratio test for this model. Finally, simulation results confirm the effectiveness of the proposed methods.