Orthogonal projection method for DOA estimation in low-altitude environment based on signal subspace

The rank-deficiency and subspace leakage caused by multipath effect are the main factors that lead to performance breakdown of direction of arrival (DOA) estimation in low-altitude environment. In this paper, we propose an orthogonal projection method based on signal subspace to overcome the negative effects of multipath. First, the signal covariance matrix is recovered to full-rank by forward and backward spatial smoothing (FBSS). Then, based on the least square technique, the signal subspace is used to establish the orthogonal projection matrix. Thereby the cross covariance matrices of signal and noise parts can be estimated and eliminated to modify the sample covariance matrix. Compared with the conventional methods that only dispose rank-deficiency, the proposed method has better performances in low-altitude environment. Besides, compared with the former orthogonal projection method based on steering matrix, this method reduces the computational complexity without iterative scheme. These conclusions are verified by simulations.