Reduced-rank covariance matrix based algorithm for DOA estimation

This paper presents a new methodology based on the reduced rank covariance matrix for direction of arrival (DOA) estimation, which can be have lower complexity compare to the classical Multiple Signal classification (MUSIC). Since the reduced rank covariance matrix contains all the information that the traditional covariance has, the performance of the proposed method is almost the same with the classical method. For comparison propose, we also compare the proposed algorithm with the Nystrom based method, it also shows the advantage of the method that has proposed in this paper.