Wideband DOA Estimation Based on Sparse Representation in 2-D Frequency Domain

In this paper, we propose a novel sparse representation (SR)-based model for wideband direction-of-arrival (DOA) estimation. Different from the classical SR-based DOA model, which only explores the sparsity of sources in spatial space, the proposed model exhibits the sparsity through the fact that any individual wideband source is characterized as a unique oblique line passing through the origin in the 2-D frequency plane of the temporal-spatial array data. To further reduce the computation complexity during the use of such sparse property, the 2-D direction information is projected into a 1-D space. Simulations show the superior performance of our approach, even in the noise corruption and less temporal samples condition.