A new phase gradient autofocus technique for high resolution image formation based on fractional lower-order statistics

This paper addresses the problem of autofocusing in Synthetic Aperture Radar (SAR) imagery by introducing techniques that achieve robust image restoration in the presence of severe heavy-tailed clutter and noise. We extend the current state-of-the-art Phase Gradient Autofocus (PGA) method by employing Fractional Lower-Order Statistics (FLOS) of the phase history data. The introduced FLOS-based PGA method mitigates the effects of impulsive additive clutter in the measurements and achieves high-resolution SAR image formation. The proposed approach is compared to conventional processing by using actual radar imagery data