A computation compression technique for SAR based on vector quantization

Abstract
A technique for data compression of synthetic-aperture radar (SAR) imagery using vector quantization (VQ) is described that provides a reduction in processing requirements for SAR focusing over traditional fast Fourier transform (FFT) implementations. The computation compression technique (CCT) used is a method of trading computations for memory. By using larger memories, the amount of computation time required to implement a function can be reduced. Results show approximately a three-fold speedup over FFT implementations using an HP 350 processor. It is noted that the CCT is much more easily parallelized than the FFT approach, which can further reduce computation times.