Decoupled DOA and Polarization Estimation Based on Block 2D Sparse Representation

A two-dimensional (2D) block sparse representation method is proposed for 2D direction of arrival (DOA) and polarization estimation by using planar sparse vector sensor array. From the second-order statistics of the array output, a virtual uniform rectangular array can be constructed, based on which a 2D sparse representation model is formulated, wherein the two overcomplete bases are each associated with a one-dimensional (1D) angle of the incident signal and matrix sizes of them are a much smaller than that in 1D sparse representation model. Then a block 2D orthogonal matching pursuit algorithm is proposed to recover the block sparse matrix from the observation data of the virtual array. The DOA and polarization parameters can be estimated from the nonzero blocks in the sparse matrix. Computer simulations show the improvements of proposed method than the existing methods in estimation performance and computational complexity.