A new approach for coherent direction-of-arrival estimation

Eigen-decomposition based methods for direction-of-arrival (DOA) estimation are known to perform poorly when the signals are coherent or highly correlated. Preprocessing techniques such as spatial smoothing and coherent source detection are commonly used to tackle the problem. But these methods often require non-trivial formation of subarrays or a high computation complexity. In this paper, we present a novel method that is capable of resolving the DOAs for coherent signals without using either source detection or spatial smoothing. Furthermore, additional number of sensors are not needed.