Direction-of-Arrival Estimation Using Eigenvectors of Covariance Matrix of Circular Array

To solve the problem that the traditional high-resolution direction-of-arrival(DOA) estimation algorithms have a large amount of computation and are not robust, the eigenvectors of noise covariance matrix of a circular array in isotropic noise field are used to redefine the eigenvectors and array manifold vectors with different orders, and the data sampling covariance matrix is reduced in dimension according to the orders. Then, DOA estimation is performed using the minimum variance distortionless response(MVDR) with the new array manifold vectors and dimension-reduced data sampling covariance matrix. Simulation results show that if there is no error, the highest-order DOA estimation result is consistent with that of the traditional MVDR method. When amplitude and phase errors exist, the more robust low-order DOA estimation results are better than that of the traditional MVDR method, which indicates that the proposed method improves the anti-error robustness, and greatly reduces the amount of computation of matrix inversion. A 12-element uniform circular array is used in the experiment conducted in the sea trail, and the results of the 2nd- and 3rd-order DOA estimation are better than that of the traditional MVDR method, verifying the effectiveness of the proposed method. This DOA estimation method may contribute to applications of circular arrays to unmanned undersea systems.