Coprime array-based DOA estimation in unknown nonuniform noise environment

In this paper, we propose direction-of-arrival (DOA) estimation techniques, respectively based on covariance matrix reconstruction and matrix completion, to achieve robust DOA estimation capability in nonuniform noise environments using coprime arrays. For the covariance matrix reconstruction-based approach, by exploring the diagonal structure of the covariance matrix of the noise, the covariance matrix of the received signal vector is reconstructed through averaging its diagonal elements. Moreover, in order to handle more sources than the number of sensors, the difference coarray of coprime arrays is utilized through the vectorization of the reconstructed covariance matrix. A compressive sensing (CS) based DOA estimator is then formulated to provide sparsity-based DOA estimation. For the matrix completion-based approach, we take the full advantage of the difference coarray lags and obtain the noise-free covariance matrix of the virtual uniform linear array by using the matrix completion technique to recover the removed diagonal elements and missing holes in the virtual array covariance matrix. Then CS-based and MUSIC-based DOA estimators are respectively designed to perform DOA estimation using the estimated noise-free covariance matrix. Simulation results verify the effectiveness of the proposed algorithms. (C) 2018 Published by Elsevier Inc.