Research On The Software Implementation Technique Of SAR Echo Simulation

Synthetic Aperture Radar(SAR)is a modern high-resolution imaging radar,which has been widely used in various fields.The simulation of SAR echo data is becoming more and more necessary for the R&D of the SAR system.SAR echo simulation is a technology that simulates the original echo data of SAR through computer simulation.It can facilitate the development of SAR imaging algorithms,the design of SAR system parameters,and the performance verification of SAR systems.The content of this thesis is as follows:1.A brief analysis of the characteristics and processing of SAR signals was made,and the principle of synthetic aperture and its high-resolution principle were introduced.This paper deduced the mathematical model of SAR echo,briefly introduced a SAR imaging algorithm-Range Doppler imaging algorithm.2.Based on the airborne SAR model,the traditional time domain SAR echo simulation algorithm was studied,and the specific implementation method was given.According to the characteristics of the echo data under some specific conditions,a matrix data migration algorithm was proposed to reduce the simulation computation;In addition,related researches on the time-frequency domain algorithm were performed and the factors influencing its computational complexity were analyzed.For the problem that the time domain algorithm and the time-frequency domain algorithm are too computationally intensive in the simulation of large scenes,two methods were introduced.The new inverse imaging echo simulation method,I?K algorithm and ICS algorithm,the paper deduced some formulae of the two methods,and analyzed their characteristics and calculations.Finally,various algorithms were used to simulate the echoes of the points,lattices,and two-dimensional targets.The differences in time-consuming and resolution differences after imaging were analyzed.3.Relevant studies on echo modeling of 3D terrain were conducted.The basic models commonly used in 3D terrain simulation,Fractal Brownian motion,was introduced.The advantages and disadvantages of common algorithms for generating fractal terrain were compared,and a certain degree of roughness mountain terrain was generated by using the diamond-square algorithm.Using high precision time domain echo algorithm for echo simulation,and the analysis of its imaging results was made.4.The parallel toolbox used to speed up simulation and solve big data operation in MATLAB software was introduced.The relevant principles of the parallelization acceleration and the main points of program modification were elaborated.the characteristics of each echo simulation algorithm were analyzed,and the suitable parallelization acceleration was given.Finally,using a distributed computing server and multiple computers to build a computer cluster,accelerated experiments of time-domain and time-frequency algorithms were performed on large-field echo simulations,and a comparative analysis was performed on the acceleration effects of different numbers of computational cores.