Weighted spatial smoothing for direction-of-arrival estimation of coherent signals

In this paper a novel and uniform frame of spatial smoothing (SS) technique for DOA estimation of coherent signals is proposed, which we call "weighted spatial smoothing (WSS)". The WSS technique takes full advantage of the auto-correlation and cross-correlation information of subarrays output and makes a weighted sum of all sub-matrices of the array covariance matrix to form an equivalent subarray covariance matrix. The optimization of weight matrix is under the constraint that the equivalent source covariance matrix is diagonal. Simulation results demonstrate that the WSS technique possesses a better resolving ability and a lower SNR resolution threshold than SS for DOA estimation of closely-spaced coherent signals, especially in the case that the size of the subarray is small.