Research On Parallel Simulation Of Jamming System Base On GPU Acceleration

The electromagnetic environment of the modern battlefield is becoming more and more complex,and the speed and efficiency of radar jamming signal and system simulation are increasing,but it still cannot meet the real-time simulation of jamming tactics and jamming strikes.It is mainly manifested in two points: one is the large number of threat radiation sources in the battlefield,and the amount of data to be modulated is huge;the other is to achieve effective interference,the signal modulation process and modulation types are complex.In order to solve the above problems,it is necessary to apply the graphics processing unit(GPU)to the radar signal processing process.At the same time,the introduction of parallel computing technology will provide strong support for the parallel modulation and processing of radar jamming signals.This thesis takes the electronic warfare engineering simulation as the background,and conducts in-depth research around the GPU-accelerated jamming system simulation technology.On the basis of relevant research on jamming signal simulation algorithms,using the large-scale parallel computing power of GPU,the GPU acceleration scheme generated by complex radar jamming signal modulation is analyzed and designed.In addition,by comparing the running time of the serial version and the parallel version of the interference signal simulation,the performance difference between the two will be analyzed.Finally,the relationship between the GPU parallel acceleration resources and the algorithm acceleration ratio will be analyzed based on the calculation results of different hardware platforms.The research work of this article mainly includes the following four parts:First,by explaining the radar jamming method and jamming signal model,analyze the necessity of jamming signal simulation and its parallelization acceleration.Aiming at the radar jamming signal simulation system,this thesis introduces the principle of convolution noise jamming and false scene jamming,analyzes the modulation process and simplified model in detail,and analyzes the key points and complex points of the serial modulation calculation in the model simulation.The necessity of parallel simulation of interference signals.Second,introduce GPU architecture and CUDA programming.This thesis introduces GPU hardware architecture and analyzes its parallel computing capabilities.On this basis,CUDA's heterogeneous programming model method and related optimization strategies are introduced.Provide a theoretical basis for the acceleration of the parallelization of interference signal simulation.Then,the parallel analysis and acceleration scheme design of the interference signal simulation are carried out.Among them,starting from the two algorithms of convolutional noise interference and false scene interference,the processing flow and time complexity are analyzed.On this basis,the optimization ideas of this thesis are proposed and the implementation scheme is designed in detail.Based on the use of the CPU+GPU heterogeneous platform to accelerate the key steps of the algorithm,and use CUDA programming to perform parallel simulation of the algorithm.Finally,the parallel acceleration implementation schemes corresponding to the two interference signals are simulated.By recording and analyzing the CPU serial operation time,GPU parallel operation time and speed-up ratio,scientifically evaluate the acceleration effect of the implementation plan,and conduct in-depth research on the relationship between parallel resources and speed-up ratio through different hardware platform simulations.The simulation experiment results show that it is feasible and effective to use GPU to accelerate the parallel simulation of radar jamming system.