Autofocus of SAR (Synthetic Aperture Radar) Imagery Degraded by Ionospheric-Induced Phase Errors

摘要

It has been suggested that synthetic aperture radar (SAR) images obtained from platforms such as SEASAT are subject to potential degradation by ionospheric-induced phase errors. This premise is based upon data from various satellite experiments that indicate large levels of phase scintillation in auroral zone data. Current models for phase errors induced by the ionosphere suggest that the phase error power spectrum is power law. This implies that the resulting phase errors contain significant components up to the Nyquist limit. Traditional sub-aperture based autofocus techniques, designed to correct uncompensated platform motion errors, are inadequate due to their inability to estimate higher order error terms. A new non-parametric phase error correction scheme developed at Sandia National Laboratories, however, has been demonstrated to remove phase errors of arbitrary structure. Consequently, our new algorithm is a viable candidate for correcting ionospheric phase errors. In this paper we show examples of SAR images degraded by simulated ionospheric phase errors. These images demonstrate that such errors cause smearing with complicated sidelobe structure. Restoration of these images via the new algorithm illustrates its superiority to classical sub-aperture based autofocus techniques. (ERA citation 14:023027)