Direction-of-Arrival Estimation Under Coexistence of both Circular and Strictly Non-Circular Signals in Coprime Linear Array

Abstract

Direction-of-arrival (DOA) in coprime linear array involves the problem that existing algorithms have less taken signal features into consideration. Considering that there is coexistence of multiple both circular and strictly non-circular signals in communication systems, an approach to DOA estimation of high accuracy and degree of freedom is proposed in this letter. We first use the ellipse covariance matrix and matrix completion method to reconstruct the covariance matrix of a virtual linear array. This virtual matrix only concludes the components of noncircular signals. Then, we similarly use the covariance matrix and matrix completion method to reconstruct a new covariance matrix including all signals of another virtual linear array. We use the virtual non-circular covariance matrix to eliminate the non-circular components in another virtual matrix. Finally, this matrix only concludes the circular signals. And we apply multiple signals classification to obtain the DOAs of two kinds of signals separately. Simulation results show that the proposed algorithm can obtain a highest degree of freedom than existing algorithms and hold on a preferable accuracy.