High-Resolution DOA Estimation Using Sparse and Low-Rank Structures for Distributed Arrays

This paper considers high-resolution direction of arrival (DOA) estimation for distributed arrays under unknown phase offsets among subarrays. We propose a structured prior that jointly exploits angular block sparsity and low-rank features across subarrays. Under the constraint of a small data fitting error, the objective function minimizes the sum of the ranks of all blocks corresponding to each candidate direction. The rank minimization is then relaxed via a sum of nuclear norms and further converted as a semidefinite program (SDP) to enable stable optimization. Simulations are conducted under both homogeneous and heterogeneous subarray configurations to compare the proposed method with three existing algorithms. Results show that the SDP-based low-rank formulation yields the sharpest spectra and robust performance with few snapshots.