Iterative Weighted Covariance Matrix Estimation Method for STAP Based on Generalized Inner Products

Accurate estimation of the clutter covariance matrix is the core issue of STAP. The sample covariance matrix estimation method based on maximum likelihood criterion is only applicable to homogeneous environment. For improving the estimation precision of covariance matrix under the heterogeneous environment, an iterative weighted covariance matrix estimation method is proposed. The proposed method determines the weighting factors of all samples according to the distance of Generalized Inner Product (GIP) value from the statistical average, and it improves the estimation precision further by establishing the probability histogram of GIP and iterative processing. The simulation results show that the proposed method can improve the performance of covariance matrix estimation under nonhomogeneous environment.