Preliminary results on subsampling effects on range migration correction in SAR imaging

Compressive sensing (CS) is becoming a popular technique for data acquisition and image reconstruction. However, sparse reconstruction often involves iterative inversion of a large system of equations. This paper explores the effects of slow-time subsampling on the Range Migration Algorithm (RMA) for Synthetic Aperture Radar (SAR). The objective is to consider methods of merging traditional image formation steps, such as range curvature correction, with CS methods for sparse reconstruction. If higher-order phase compensations can still be performed via existing algorithms, then it may be possible to perform image formation via a parallel set of reduced-size sparse reconstructions rather than one large reconstruction. We study range migration correction and evaluate the imaging function for scatterers distributed across the field of view and for various compression levels.