Integrated Compressive Sensing and Autofocus for SAR Imaging

High-resolution synthetic aperture radar (SAR) imaging requires large bandwidth and storage capacity, and trajectory errors caused by the motion of airborne or unmanned aerial vehicle platforms can degrade image quality. To address these challenges, compressive sensing (CS) is introduced to reduce the SAR raw data volume, and a novel method is proposed for SAR image formation that combines CS-based recovery with an autofocus technique. For CS-based reconstruction, a modified low-rank and sparse decomposition (MLRSD) scheme is employed based on the range-Doppler algorithm, and an autofocus algorithm is subsequently applied to mitigate image quality degradation caused by residual phase errors after motion compensation. Through simulations using SAR modeling data, we present the convergence characteristics of several autofocus algorithms combined with the MLRSD-based reconstruction technique. The quality of the recovered SAR images was compared, and the best autofocus algorithm was selected for the proposed method. Finally, the proposed method was applied to real vehicular SAR measurement data to perform SAR image reconstruction and autofocus operations using only a part of the raw data. The experimental results demonstrate that the SAR image reconstructed using the proposed method exhibits superior quality compared with those obtained using conventional techniques.