Enhanced compressed sensing autofocus for high-resolution airborne synthetic aperture radar

Abstract

The limited accuracy of the navigation system leads to deviations between the real and measured trajectories in airborne synthetic aperture radar (SAR) which results in residual motion (phase) errors and seriously degrades the image quality. The residual errors are usually corrected by using autofocus methods. In this paper, we investigate the autofocus for a high-resolution airborne SAR system which employs stepped-frequency chirp (SFC) signal to achieve high range resolution. As a promising waveform to improve radar range resolution, the SFC signal is widely realized in practical systems. We characterize the backscattering from the scene, phase errors, and receiver noise with a linear matrix-based formulation. In addition, we propose a compressed sampling method that is directly applied to the scene and received signal to provide a sparse raw dataset. We describe the autofocus problem as a constrained optimization problem and propose a new compressed sensing (CS)-based autofocus method to solve it. This method iteratively estimates the phase error and reconstructs the high-resolution SAR image. The image reconstruction is based on the proximity algorithm which applies the shrinkage proximal operator to solve the problem in a computationally efficient manner. We validate the effectiveness of the proposed method using both simulated and real SAR images. Based on visual results and a variety of metrics, the proposed method outperforms the classical autofocus approaches. Also, it provides better or comparable results, with much lower computational complexity, in comparison to that of the recent CS-based autofocus methods.