Low Complexity Geometrical Autofocusing Based on Subsequent Sub-Apertures Calibration

Synthetic Aperture Radar (SAR) is an imaging technique capable of generating high-resolution images of the observed scene. Accurate knowledge of the platform's trajectory is crucial for achieving highly focused images. Failure to meet this requirement will result in phase errors, leading to blurring effects in the final image. This work presents a low-complexity autofo-cusing algorithm based on the geometrical compensation of phase errors. The total platform trajectory is divided into adjacent, non-overlapping sub-apertures, reducing phase error contribution and enhancing the effectiveness of calibration. The correction is propagated to adjacent sub-apertures, leveraging the targets' estimated position in the first calibrated sub-aperture. After each sub-aperture undergoes geometrical compensation, they are all coherently merged, enabling the retrieval of the unblurred, full-resolution SAR image. The proposed method outperforms in scenarios with significant variations in the incidence angle, a key consideration for the implementation on low-altitude applications as UAV-borne SAR, and/or cases where the trajectory usually extends several times the synthetic aperture length, typical of stripmap mode, effectively overcoming the limitations of the PGA (Phase Gradient Autofocus) algorithm, traditionally invalidated under these conditions. In this paper, the algorithm is accurately described and validated using a simulated data set. Finally, the effectiveness of the approach is proved using real radar data.