Research On Ground-based SAR Real-time Imaging Technology Based On Multi-core DSP

With the characteristic of all-time,all-weather,simply structure,lower peak power and strong real time performance,the synthetic aperture radar(SAR)of frequency modulation continuous waves(FMCW)is used to detect wide surface of the mountain slope in the monitoring system.The real time imaging technology of ground-based synthetic aperture radar has been researched in this thesis.The summarizations of the work are as following,1.Based on the characteristic of FMCW ground-based SAR,the algorithm of real time imaging processing is researched and the module of echo signal is created for the single point target in the area of imaging.For the imaging requirement of real time performance and precise in the ground-based SAR system,the polar back projection algorithm is applied to imaging processing and the realization procession of which is improved for the purpose of real time imaging on the basis of algorithm procession and realization characteristic in digital signal processor.2.The technology of real time imaging processing is researched.Firstly,the software framework of real time processing of ground-based SAR based on multi-core DSP is designed.According to the principle of task assignment and distribution,the technology of parallel processing of ground-based SAR imaging based on multi-core DSP and the technology of input-output streams and parallel imaging processing are designed.The key technology of real time processing is studied,such as long points FFT,SYS/BIOS,multicore synchronization and communication.Finally,the correctness and effectiveness of software framework are verified based on the comparison of imaging results of measured data in MATLAB and software framework.3.Aiming to the question of real time imaging processing in big wide ground-based SAR and the limited memory resources in DSP,the image mosaic technology of real time imaging processing of big wide ground-based SAR is researched.Finally,the correctness and effectiveness of software framework are verified based on the comparison of imaging results of measured data in MATLAB and software framework.