A new approach to moving targets and background separation in multi-channel SAR

This paper describes a new approach for decoupling the slowly-moving targets and the clutter under multi-channel SAR (Synthetic Aperture Radar) scenarios. Given the phase histories received by different channels, the high-level data structures, e.g. the sparsities of moving targets, enable us to enforce advanced constraints on an optimisation problem to distinguish the dynamic targets from static background. Specifically, an iterative method is employed for the decomposition by alternatively updating the arguments. The dynamic parts can then be used for the subsequent analysis on states of the targets, and the static parts, in the case that the background does not change significantly over time, can be inherited in subsequent sub-aperture processings. The presented approach is demonstrated on simulated three-channel SAR data.