Research Of SAR Echo Simulation Based On GPU And FPGA

Synthetic Aperture Radar(SAR)with the outstanding features of all-weather,fullweather,long-range to provide high-resolution images,has been widely used in military and civilian aspects.The simulation of the echo signal and the processing of the radar signal are the basis of the synthetic aperture radar system and its imaging algorithm.As the calculation of the radar signal is too large,there is a long time-consuming,low efficiency,high cost and other shortcomings to use the traditional computer processor processing.In order to solve the above problems,this thesis studies the parallel SAR echo simulation technology by using the parallel computing capability of graphics processing unit(GPU)and field programmable gate array(FPGA),and improves the efficiency of SAR echo simulation.Based on the basic principle of SAR and its imaging algorithm,this thesis designs a SAR echo simulation platform based on GPU and FPGA,which includes SAR echo simulation based on GPU and SAR baseband signal processing based on FPGA.The major devotement and innovation of this thesis is as follows:Firstly,explain the basic principle of synthetic aperture radar and the distance Doppler imaging algorithm,analyze the basic principle of the LFM signal and the pulse compression,and compare the traditional SAR echo simulation technology.Secondly,design a CUDA parallel program based on GPU,build the SAR echo simulation model of airborne side-looking mode.Focus on the design of the parallel algorithm in SAR echo simulation,Based on the characteristics of powerful floating point computing and parallel computing of GPU,the echo simulation algorithm of point target and surface target is realized on the GPU,and its simulation time is analyzed.Finally,design the SAR baseband signal processing system based on FPGA with the parallel processing capability of FPGA and the system level modeling tool of System Generator,which includes the key technologies of radar signal processing,such as parallel LFM signal,quadrature mixing,filter extraction and convolution.Through the debugging of the key interface of the system,the function of the platform is verified.