Sparse SAR Imaging Based on Periodic Block Sampling Data

Recently, a novel design scheme of low-earth-orbit spaceborne mini-synthetic aperture radar (MiniSAR) system is proposed to exploit the integrated transceiver to collect the azimuth periodic block sampling data by using alternated transmitting and receiving operations. Because such collected data are downsampled, the images recovered by the typical matched filtering (MF)-based methods have the problems of obvious azimuth ambiguities, ghosts, and energy dispersion. To find a suitable method for such data, with the help of sparse signal processing technique, we first introduce sparse synthetic aperture radar (SAR) imaging with $\ell _{1}$ -norm regularization-based approximated observation method to recover the large-scale considered scene. To further improve the imaging performance, a novel approximated observation unambiguous sparse SAR imaging method via $\ell _{2,1}$ -norm is proposed. Compared with $\ell _{1}$ -norm -based method, the recovered image by the proposed one achieves better imaging quality with reduced azimuth ambiguities and ghosts. Experimental results on simulated and real data validate the proposed method.