关键词: 阵列信号处理, 波达方向, 压缩感知, 稀疏重构, 高分辨, 宽带, 谱估计

Abstract: With the utilization of spatial sparseness of wideband sources, a wideband direction of arrival (DOA) estimation problem can be translated into a sparse reconstruction problem, based on which a novel wideband DOA estimation algorithm is presented. It decomposes a wideband signal into multiple sub-band signals, and utilizes jointly the common spatial sparse pattern of these sub-band signals. The proposed algorithm can be viewed as an extension of the original sparse Bayesian learning method to the case of multiple redundant dictionaries. Additionally, by the singular value decomposition (SVD) performed on the multiple snapshots of the array received signal, the signal subspace is extracted as the input of the algorithm, which effectively reduces the computational complexity and simultaneously improves the robustness to noises. Compared with classical wideband DOA estimation methods, this algorithm performs better even in the cases of low signal-to-noise ratio, limited available snapshots and high correlations of sources. Its performance is insensitive to a biased estimate of source number. Simulation results verify its performance advantages over existing subspace-based wideband DOA estimation methods.

Key words: array signal processing, direction of arrival, compressed sensing, sparse reconstruction, high resolution, wideband, spectral estimation