2D-DOA and Polarization Estimation Using a Novel Sparse Representation of Covariance Matrix With COLD Array

In this paper, we propose a novel sparse signal representation (SSR)-based algorithm called the
$\ell _{1}$
-PSRCM with cocentered orthogonal loop and dipole (COLD) array to estimate two-dimensional (2D) direction of arrival (DOA) and polarization parameters. Considering the characteristics of polarization sensors, a polarized sparse representation model of covariance matrix is constructed, whose overcomplete dictionary and sparse coefficient matrix only depend on DOA and polarization parameters, respectively. In so doing, the proposed algorithm can make full use of the spatial and polarized information contained in the received data, thereby improving the estimation accuracy. In addition, to reduce the computational complexity and suppress the effect of noise, the modified
$\ell _{1}$
-PSRCM algorithm with the real-valued sparse coefficient matrix and noise-free sparse representation model is proposed. Finally, we present the two-dimensional multiresolution grid refinement (2D-MGR) method to reduce the heavy computation burden when the spatial grid is dense. Simulation results validate the superiority of the proposed algorithms.