Research On ISAR Real-Time Imaging System Based On GPU Parallel Acceleration

Inverse Synthetic Aperture Radar(ISAR)can not only obtain high-resolution images of the observed target,but also obtain data such as its structure and size,providing the basis for subsequent target identification and classification.ISAR imaging simulation refers to obtaining echoes consistent with the actual radar observation targets through simulation based on the simulation platform,and then processing it with imaging algorithm to obtain the image information of the target.With the passage of time,people have placed higher demands on the quality and resolution of ISAR imaging.The improvement of ISAR imaging resolution is generally achieved by increasing the bandwidth of the radar transmission signals and the synthetic aperture time.These methods will result in a dramatic increase in radar echo data,reducing signal processing speed and ISAR imaging efficiency.Compared with the Central Processing Unit(CPU),the Graphics Processor Unit(GPU)has a multi-threading mechanism and is highly parallel,so it has superior parallel processing performance.This thesis applies the GPU Compute Unified Device Architecture(CUDA)to ISAR echo simulation and imaging processing,and combines Open GL and Qt technologies to design and develop an ISAR real-time imaging system with user-friendly human-computer interaction,high imaging simulation performance and network communication capabilities.The main research contents of the thesis are as follows:1.Aiming at the system platform design,the block diagram and software composition of the ISAR real-time imaging simulation system are analyzed,and the development technologies including GPU general computing model,Qt framework,Open GL,and TCP/IP network communication are investigated to provide technical support for subsequent software development.2.For the system algorithm design,the classical methods of electromagnetic calculation,ISAR echo model,ISAR imaging and motion compensation are investigated.Firstly,the advantages and disadvantages of the commonly used electromagnetic calculation methods are analyzed,and the bouncing ray method is selected as the electromagnetic simulation algorithm of this system.Then,based on the turntable model,the echo signal of the ISAR target is derived,and the basic principles of ISAR imaging and motion compensation techniques are introduced in detail.Finally,the feasibility of the motion compensation algorithm is verified using point targets and Yak-42 aircraft real measurement data,which provides theoretical support for the design of the GPU parallel scheme for the subsequent ISAR imaging simulation.3.To address the problems of low computing efficiency and real-time performance in the ISAR imaging simulation process,CUDA technology,the GPU general computing architecture,is applied to the ISAR imaging system,and the high parallelism of CUDA is used to accelerate the ISAR echo simulation and imaging processing in parallel to improve the real-time performance of ISAR imaging.4.For the design of the human-computer interaction interface of the imaging system,the three-dimensional model display,network communication,and human-computer interaction modules are analyzed and designed based on the Qt modular idea so as to meet the requirements of the simulation software to control the GPU for ISAR imaging through human-computer interaction.