Research Of Parallel SHARC System In Application Of SAR Signal Processing

Signal processor is the key part of synthetic aperture radar(SAR). The development of SAR relates to the level of signal processing devices to a great extent. Nowadays, high capability DSPs often appear, that makes easy realization of many complex theories and algorithms on hardware, and promotes the development of SAR technology. Parallel processing system consisted of ADSP21060-SHARC parallel 32-bit floating point DSP, produced by Analog Devices Inc, can satisfy signal processing requirement in SAR application fields.This dissertation analyzes Phase Gradient Autofocusing (PGA) algorithm and speckle reduction algorithm with Lee filter. And the two algorithms are realized on the SHARC parallel processing system.Chapter 1 analyzes phase error in SAR, autofoucusing methods, especially PGA algorithm, and the realization method of PGA algorithm. Chapter 2 introduces the characteristics and mechanism of speckle and the technology of speckle reduction in which Lee filter algorithm is introduced mainly.The first two chapters are theoretical base of realization with hardware.Chapter 3 presents the interior architecture of SHARC processor and the parallel system consisted of multi-processor.Chapter 4 introduces Darlington PCI board software and hardware developing environment and debugging steps.Chapter 5 accounts for the hardware realization methods of PGA and Lee filter on parallel processing system, discusses some problems to be considered, gives the improving methods and presents the processing results.In the end, the dissertation summarizes the whole work, and points out the problems to be solved and the way to continue researching.