Direction-of-arrival estimation for coherent signals through covariance-based grid free compressive sensing

Abstract

This letter solves direction-of-arrival (DOA) estimation of coherent signals for uniform linear array. The proposed method reformulates the covariance matrix as a multiple-snapshot array output model to obtain a high array signal-to-noise ratio. Then the scheme of the atomic norm minimization [Yang and Xie, IEEE Trans. Signal Process. 64(19), 5145-5157 (2016)] is extended to this model, thus estimating the DOAs in a continuous space. The proposed method avoids the condition where DOAs deviate from the discretized grid, i.e., basis mismatch. The simulation and experimental results verify the good performance of the proposed method for coherent signals.