Sparse Decomposition-Based DOA Estimation Algorithm

This paper presents a sparse decomposition-based DOA estimation method derived from the Orthogonal Matching Pursuit (OMP) algorithm. The proposed approach employs matching pursuit selection by leveraging the uncorrelation between array manifold vectors, systematically identifying the manifold points that best approximate the actual source locations. Key implementation aspects include: constructing a dictionary matrix from array manifold vectors, implementing iterative correlation calculations for atom selection, and performing residual updates through orthogonal projection. The algorithm demonstrates enhanced DOA estimation accuracy through optimized sparse signal reconstruction. Furthermore, this methodology exhibits versatility across various signal processing domains, offering broad applicability and practical utility. We anticipate this approach will provide researchers in signal processing with novel perspectives and inspiration for advanced direction-finding applications.