Kronecker Product PCA for structured covariance matrix of airborne radar STAP

Abstract

This paper considers the estimation of structured clutter-plus-noise covariance matrix (CNCM) in space-time adaptive processing (STAP) for airborne radar systems. Specially, the CNCM is modeled as a sum of Kronecker products involving two lower dimensional temporal and spatial covariance matrices, with persymmetric structure. Then, resorting to the Kronecker Product principal component analysis (KronPCA) based algorithm, a novel estimator of the high dimensional and persymmetric CNCM is proposed. Furthermore, the proposed method explores the sparse factors of the CNCM and recovers low-rank persymmetric covariance matrices. At analysis stage, we assess the performance of the proposed algorithm through simulations.