SAR Image Formation via Subapertures and 2D Backprojection

Abstract

Radar imaging requires the use of wide bandwidth and a long coherent processing interval, resulting in range and Doppler migration throughout the observation period. This migration must be compensated in order to properly image a scene of interest at full resolution and there are many available algorithms having various strengths and weaknesses. Here, a subaperture-based imaging algorithm is proposed, which first forms range-Doppler (RD) images from slow-time sub-intervals, and then coherently integrates over the resulting coarse-resolution RD maps to produce a full resolution SAR image. A two-dimensional backprojection-style approach is used to perform distortion-free integration of these RD maps. This technique benefits from many of the same benefits as traditional backprojection; however, the architecture of the algorithm is chosen such that several steps are shared with typical target detection algorithms. These steps are chosen such that no compromises need to be made to data quality, allowing for high quality imaging while also preserving data for implementation of detection algorithms. Additionally, the algorithm benefits from computational savings that make it an excellent imaging algorithm for implementation in a simultaneous SAR-GMTI architecture.