On the Role of Scene Coordinate System in Focusing of GEO SAR Data With Fast Time-Domain Algorithm

Geosynchronous-Earth-orbit (GEO) synthetic aperture radar (SAR) is able to achieve targeted observation, thanks to its flexibility in antenna steering, while focusing GEO SAR data is challenging due to the highly dynamic imaging geometry. Its processing algorithm is required to be able to deal with different view angles at different orbit segments, and the non-Euclidean scene geometry must be considered in the algorithm model and processing chain. This article focuses on developing the processing algorithm applicable to the GEO SAR, especially at high squint. The analytic expressions of the squint-mode wavenumber support and point spreading function are derived. Then, we propose a generalized fast Cartesian factorized back-projection algorithm, which can deal with high-squint data, and the scene coordinate system can be built with high flexibility. Especially, the role of the scene coordinate system in achieving high imaging performance is studied, and the optimal scene coordinate system is exploited. Finally, the processing results of simulated point target data are presented to validate the proposed algorithm.