Dual-Band Radar Forward-Looking Imaging Real-Time Processing Technology

When the missile strikes ground and sea targets in the dive attack section,it needs to carry out active imaging homing terminal guidance on the forward-looking area under the interference of complex electromagnetic environment.However,for the imaging of the forward-looking area,the traditional SAR system can only use the arc-shaped trajectory for squint imaging due to insufficient Doppler bandwidth,which increases the target strike time.Traditional single-frequency radars often make wrong judgments due to interference when working,while dual-frequency radars have strong anti-jamming capabilities through the combination of high and low frequency bands.At the same time,the missile-borne radar platform moves fast and has high real-time imaging performance,so the research on realtime processing technology of dual-frequency radar forward-looking imaging is of great significance.Based on the background of a certain project,this paper concentrates two forward-looking algorithms working in two frequency bands in the same radar imaging system,and focuses on the parallel architecture design and algorithm engineering implementation of the DSP part of the system.This paper mainly starts from five aspects:the principle of dual-frequency radar forward-looking imaging,software requirement analysis,system process design,parallel architecture design and parallel implementation of algorithms,and completes the real-time forward-looking imaging of dual-frequency radar system.1.The principle of single-pulse forward-looking imaging and microwave correlation forward-looking imaging are introduced.The beam scanning technology is used to effectively improve the angular flicker problem in the traditional single-pulse forwardlooking imaging technology,and the OMP algorithm is used to effectively solve the problem in correlation imaging.Compressed sensing model engineering and real-time processing problems.In order to make reasonable use of hardware resources,by analyzing the characteristics of each step of the algorithm,the characteristics of FPGA and DSP chips,the mapping of the algorithm in hardware is completed,and the system workflow is designed at the same time,and the algorithm switching strategy is analyzed,which lays the foundation for the subsequent parallel architecture design.2.In order to solve the problems of large amount of echo data,complex imaging algorithm,and cumbersome process in the radar system,so that the system can meet the real-time requirements,according to the characteristics of the forward-looking algorithm,four aspects of memory management,high-speed interface,parallel processing,and multi-core synchronization are implemented.Starting from this aspect,a DSP multi-core parallel architecture is proposed.Aiming at the problem of large amount of data in the process of data processing,the memory space is allocated reasonably,and the memory bandwidth is utilized to the maximum extent.Aiming at the high-speed data interaction between FPGA and DSP in the system,the high-speed interface is designed,and the data transmission rate test is carried out.By analyzing the characteristics of each step in the imaging algorithm,the parallel processing method of each step in the algorithm is designed,and completed multicore synchronization.At the same time,the cache coherence problem,big data moving problem,matrix transposing problem and compressed sensing model engineering problem in engineering are analyzed,and corresponding solutions are given.3.According to the designed multi-core parallel architecture and algorithm principle,the tasks of each core are allocated,and then a beam scanning-based monopulse forward-looking imaging technology and compressed sensing-based microwave correlation are implemented on the DSP side of the system.Forward-looking 3D imaging technology.At the same time,in order to solve the problems of low efficiency and large amount of algorithm calculation when the algorithm is engineered using C language,the program is deeply optimized from the perspectives of compiler,memory,code and algorithm.Finally,a test platform was built,and the stability,real-time and correctness of the system were verified by means of semiphysical simulation through data acquisition and playback.