An efficient algorithm for 3-D SAR imaging

The RELAX algorithm is a successful solution to Elie side-lobe leakage problem in generation of three-dimensional (3-D) synthetic aperture radar (SAR) images from sparse frequency space data collected via curved synthetic aperture. RELAX utilizes the 3-D FFT to obtain the 3-D position estimate of scatterer at each iteration. Since the data processed by 3-D FFT contains a large number of zeros, which introduce huge additional computations, the RELAX algorithm is not efficient. In this paper, a similarly structured algorithm, which is referred to as lower dimensional position searching (LODIPS) algorithm, is proposed for 3-D SAR imaging from the curved synthetic aperture. Due to the utilization of lower dimensional FFTs, the LODIPS algorithm is more efficiency. Simulation results show that the estimates obtained via both RELAX and LODIPS can achieve the corresponding Cramer-Rao bounds (CRBs) at low SNR.