Fast Fourier-Domain Optimization Using Hybrid L-1-/L-p-Norm for Autofocus in Airborne SAR Imaging

Abstract

In maneuverable airborne synthetic aperture radar (SAR), high-frequency phase errors (HPEs), which are hard to be compensated by the phase gradient autofocus, often occur causing spurious targets and contrast degradation in SAR imagery. The metric-optimization approaches improve the phase error estimation scalability, including HPEs, but show much lower speed because of the computational burden by the inverse fast Fourier transform (IFFT). Proposed autofocus based on the metric optimization enables the rapid estimation with the scalability including the HPE by eliminating the IFFT. In addition, the metric is adapted to the range compressed data corresponding to the input of the autofocus taking advantage of the convergence speed and the stability in its optimization using a hybrid form of L-1- and L-p-norm. In the simulation, we focused on the suggestion of the proof of performance generality.