Research On SAR Echo Simulation And Imaging Algorithm Baesd On GPU And CUDA

Synthetic Aperture Radar(SAR)has prominent features of providing high-resolution images throughout the day,all-weather,and long-distance,and is widely used in military and civilian applications.The simulation of echo signals and the processing of radar signals are the basis for the study of synthetic aperture radar systems and their imaging algorithms.For radar signal processing,the computational complexity is too large.When processing with a traditional computer processor,there are disadvantages such as long time-consuming,inefficient,and high cost.For the existing problems,this paper makes use of the parallel computing capability of GPU and MATLAB parallel computing toolbox as well as MATLAB's powerful numerical analysis and matrix computing capabilities to deeply study the parallel SAR echo simulation technology.And call CUDA(Compute Unified Device Architecture)library technology in MATLAB to improve the calculation efficiency of SAR echo simulation.This paper mainly studies the basic principles of SAR,Range Doppler imaging algorithm(RD algorithm),Chirp Scaling algorithm(CS algorithm)and OSS algorithm,designs and optimizes algorithms,and uses GPU to increase the speed of the algorithm.Firstly,the basic principle of Synthetic Aperture Radar(SAR)and Range Doppler imaging algorithm are introduced.The basic principles of LFM signal and pulse compression are analyzed in detail.The traditional SAR echo simulation technology is analyzed and compared.Second,designed a GPU-based CUDA parallel program.And established SAR echo signal simulation model for strip mapping under airborne positive side conditions.Using GPU's powerful floating-point computing capabilities and parallel computing capabilities,the echo simulation algorithms for point targets and surface targets were implemented on the GPU,and their simulation time was analyzed.Finally,combining the OSS algorithm,the MATLAB parallel computing toolbox and the CUDA algorithm library are combined to optimize the program.The conclusion is drawn through simulation imaging and time comparison.