PRECISION SAR PROCESSING USING CHIRP SCALING

A space-variant interpolation is required to compensate for the migrat ion of signal energy through range resolution cells when processing sy nthetic aperture radar (SAR) data, using either the classical range/Do ppler (R/D) algorithm or related frequency domain techniques. In gener al, interpolation requires significant computation time, and leads to loss of image quality, especially in the complex image. The new chirp scaling algorithm avoids interpolation, yet performs range cell migrat ion correction accurately. The algorithm requires only complex multipl ies and Fourier transforms to implement, is inherently phase preservin g, and is suitable for wide-swath, large-beamwidth, and large-squint a pplications. This paper describes the chirp scaling algorithm, summari zes simulation results, presents imagery processed with the algorithm, and reviews quantitative measures of its performance. Based on quanti tative comparison, the chirp scaling algorithm provides image quality equal to or better than the precision range/Doppler processor. Over th e range of parameters tested, image quality results approach the theor etical limit, as defined by the system bandwidth.