A Hybrid-Domain Algorithm for High-Speed High-Squint SAR Imaging With Curved Trajectory via Fifth-Order FNCS Processing

Because of the severe 2D range-azimuth coupling (RAC) and Doppler space-variance (DSV), high-speed high-squint (HSHS) synthetic aperture radar (SAR) imaging with curved trajectory is still extremely problematic. To address this issue, a hybrid domain algorithm via fifth-order frequency nonlinear chirp scaling (FNCS) processing is proposed in this paper. The proposed algorithm comprises three modules: range focusing, azimuth compression, and geometric correction (GC). To enhance model accuracy while simplifying the coefficient representation, a fifth-order slant range model based on equivalent velocity and parameter vectorization is initially established. After the pre-processing employed to remove bulk RAC, range focusing is accomplished using a range-variant Stolt interpolation. Then, to eliminate DSV for azimuth compression, a fifth-order FNCS processing is presented based on a minimalist derivation approach. This approach derives an explicit expression for the FNCS-processed signal directly from the determined model and NCS orders to aid in the solution of undetermined parameters. Subsequently, an instantaneous projection geometry is proposed for GC, with the capability to eliminate arbitrary geometric distortions. Finally, the results of processing simulation and real SAR data demonstrate the validity and superiority of the proposed algorithm.