Parameter-searched 2D orthogonal matching pursuit algorithm for sparse GB-SAR imaging

For frequency-modulated continuous-wave (FMCW) ground-based synthetic aperture radar (GB-SAR) imaging, classical matched filtering-based algorithms suffer from limited resolution and high-level sidelobes. By exploiting the sparsity of targets, compressive sensing (CS) based algorithm can achieve higher resolution and reduced sidelobe level. Conventional CS based imaging algorithm is a typical sparse vector recovery problem that can be solved by greedy algorithms. Three problems, e.g., computational complexity, memory usage, and reconstruction accuracy, will appear when conventional greedy algorithms are applied to sparse GB-SAR imaging with large illuminated area, high resolution requirement, and off-grid targets. In this paper, we proposed a new greedy algorithm, i.e., 2D parameter-searched orthogonal matching pursuit (2D POMP) algorithm, to solve these problems. The imaging problem is modeled as a sparse matrix recovery problem, which can be solve by the proposed algorithm effectively and accurately. Compared with the imaging algorithms based on OMP, the computational burden and memory usage of the proposed algorithm is decreased, while the reconstruction accuracy is improved. Simulation results validate the proposed imaging algorithm.