One-Bit DOA Estimation of Incoherently Distributed Sources via Sparse Covariance Fitting in Massive MIMO Receive Array

To address the challenges posed by the high cost and power consumption of massive multiple-input multiple-output receive array, as well as the limitations of point-source-based direction-of-arrival (DOA) estimation methods when applied to incoherently distributed (ID) sources, a one-bit DOA estimation approach tailored for ID sources is proposed in this work. First, we formulate the covariance fitting equation using the correlation between one-bit covariance matrix and its normalized unquantized counterpart. Due to its nonconvexity, we discretize the spatial domain and transform it into a convex second-order cone programming problem. Then, we derive a cyclic minimization algorithm that iteratively minimizes the resulting cost function in closed forms. Using a first-order Taylor expansion, we reconstruct the noiseless signal covariance matrix, thereby accurately estimating the angular spread. Furthermore, we derive the approximate Cramér–Rao bound. The superiority in terms of accuracy and efficiency is demonstrated through simulation results.