A precision chirp scaling SAR processor extension to sub-aperture implementation on massively parallel supercomputers

A new concept in SAR raw data focusing algorithms is discussed. The so called “chirp scaling” (CS) technique has allowed the complete elimination of the interpolation step required in conventional Ω-K wave domain processing algorithms. This drives to high performance the implementation of aberrationless 2D processors, both for SAR focusing and for analogue problems (seismic wave migration, tomography, etc.). Furthermore the suppression of Stolt interpolation leads to the possibility of a very efficient implementation on massive parallel supercomputers. The aim of the paper is not intended to disclose new theoretical secrets of the SAR processing algorithms, but simply shows an example of application of CS on a parallel supercomputer, indicating the problems met in the course of the author's experience