Knowledge-aided shrinkage interference covariance matrix estimation in STAP

Abstract

This paper proposes a knowledge-aided shrinkage interference covariance matrix estimation approach in space time adaptive processing (STAP) which exploits the prior interference covariance matrix based on clutter model as the shrinkage target matrix. The oracle shrinkage coefficient under Gaussian assumption is derived and the estimated consistent shrinkage coefficient is also provided since the true interference covariance matrix in the oracle coefficient expression is usually unknown. With the comparison to the identity matrix and scaled identity matrix which are most commonly used as the shrinkage target matrix, the superiority of using prior knowledge is validated using simulated data. To evaluate the sensibility of the proposed method to the prior knowledge error, performance employing inaccurate knowledge is also considered in this paper. Simulation results show that the proposed method can significantly improve the STAP performance, even when the prior knowledge has some errors.