A Low-Complexity DOA Estimation Algorithm for Distributed Non-Uniform Arrays

Distributed uniform arrays require both uniform subarrays and precise alignment along a common baseline. However, in practical deployment, subarray motion and structural variations can lead to deviations from the intended configuration, forming a distributed non-uniform array. In this case, the sampling matrix of the distributed array may not be as expected, and it is necessary to reacquire the global sampling matrix. Additionally, measurement errors in the subarrays further contribute to deviations from the intended uniform structure, which complicates Direction-of-arrival (DOA) estimation. To address these challenges, this paper derives a general sampling matrix formulation for distributed non-uniform arrays based on the subarray sampling matrices and position information. Based on Multiple Signal Classification (MUSIC) and Manifold Separation Technique (MST), we propose a distributed MST-MUSIC algorithm that reduces the computational complexity of MUSIC while maintaining angle estimation accuracy.