AgileSAR: Achieving Wide-Swath Spaceborne SAR Based on Time-Space Sampling

Abstract

High resolution and wide swath, which are related to imaging quality and observation efficiency, are the key specifications for spaceborne synthetic aperture radar (SAR). Owing to the restrictions of the Nyquist sampling theorem, it is difficult to improve both specifications simultaneously. The increase of the swath width often leads to the decrease of the spatial resolution, e.g., in scanning SAR and terrain observation with progressive scan SAR. For a sparse scene, an image containing only a few targets has massive data but little useful information. This paper proposes a novel SAR observation mode, AgileSAR, which is based on the time–space sampling method and can overcome the limitations of the Nyquist theorem. It also increases the swath width while preserving the resolution of the sparse scene. AgileSAR steers the antenna beam towards a different sub-swath, generally after one or two pulse intervals, and the average pulse repetition rate corresponding to every sub-swath is much lower than that determined by the Nyquist theorem. Compared with Sentinel-1, which can achieve 5-m resolution and 80-km swath, a single azimuth-channel AgileSAR system can achieve 5-m resolution and 300-km swath for a sparse scene, once the corresponding system parameters are designed. The $\text{l}_{1}$ relaxation method is used to reconstruct sparse SAR images, and the reconstruction performance is quantitatively analyzed based on the estimation error. The simulation results validating the proposed method with sub-Nyquist samples can achieve approximately similar performance as conventional SAR with Nyquist samples.