Low-Complexity DOA Estimation for Coherently Distributed Sources With Gain-Phase Errors in Massive MIMO Systems

In this work, a low-complexity direction of arrival (DOA) estimation method is proposed for coherently distributed (CD) sources considering gain-phase errors in a massive multiple-input multiple-output (MIMO) system. To alleviate the computational burden of calculating the covariance matrix and its eigen decomposition in massive MIMO systems, a Nyström-based signal subspace estimation technique is proposed, which only requires the computation of two submatrices of the covariance matrix. Furthermore, the DOAs, angular spreads and gain-phase coefficients are obtained in closed form via array shift invariance, provided that the partly antennas are well-calibrated. In addition, we derive the Cramér-Rao bound for the discussed problem. Finally, the proposed method performs attractive effectiveness and robustness compared to other methods through numerical simulations.