Sparsity-Driven Stripmap SAR Imaging and Phase Error Estimation Based on Phase Curvature Autofocus

The phase curvature autofocus (PCA) is an autofocusing algorithm with high efficiency and excellent robustness. However, its accuracy will be severely affected by the signal-to-noise ratio (SNR) of range-compressed data, and it is not suitable for downsampled data. In this letter, we propose a novel sparse autofocusing imaging method for stripmap synthetic aperture radar (SAR) phase error correction and image reconstruction. In the proposed method, the initial SAR image is first recovered by the Bi iterative soft thresholding (BiIST) algorithm, which is a sparse imaging technology that can suppress noise and maintain image phase information. Then, PCA is used to process the reconstructed image with phase information to estimate the phase error. Finally, the approximated observation-based sparse imaging method is used to obtain the precisely focused SAR image from the compensated full-sampled or downsampled echo data. Experimental results based on simulated data with adding different kinds of phase error show that compared with the reconstructed results of chirp scaling algorithm (CSA), traditional PCA, and iterative soft thresholding (IST)-based sparse imaging method, the contrast of the proposed method-based image can be improved at least 30%.