Design And Implementation Of Real-time Target Detection Algorithm For Millimeter Wave MIMO Radar Based On ZYNQ Platform

Millimeter-wave Multiple Input Multiple Output(MIMO)radar,equipped with multiple transmitting and receiving antennas,utilizes the separation of electromagnetic waves emitted by different transmitting antennas to form a large aperture virtual array,thereby achieving higher target angle estimation performance.Millimeter-wave MIMO radar also offers advantages such as strong penetration,wide bandwidth,and small antenna size,resulting in shorter acquisition time for a frame of echo data,higher target detection resolution.However,millimeter-wave MIMO radar generates a higher volume of data,increasing the processing time of the target detection system.The ZYNQ platform integrates a Processing System(PS)and Programmable Logic(PL),providing abundant internal hardware resources for extensive data computations.Therefore,this paper conducts research on the hardware design and implementation of real-time target detection algorithms for millimeter-wave MIMO radar.The entire process of radar data reception,real-time target detection algorithms,and result visualization is implemented on the ZYNQ-7020 platform.The data format output by the radar hardware module is analyzed,optimized,and accelerated.Hardware implementations are applied to the real-time target detection algorithm for millimeterwave MIMO radar.The main objectives of this paper are as follows:Firstly,starting from the theory of Frequency-Modulated Continuous Wave(FMCW)MIMO radar target detection,a millimeter-wave FMCW MIMO radar system model is constructed.The principles of range measurement for FMCW radar and angle measurement for MIMO radar are analyzed.The signal model for millimeter-wave FMCW MIMO radar and the principles of Digital Beam Forming(DBF)target detection algorithm are derived,followed by simulation analysis of the DBF algorithm.Next,based on the requirements of real-time target detection,and considering the characteristics of the MIMO radar hardware module and the ZYNQ-7020,a system for implementing real-time target detection algorithms is designed.By allocating tasks to the ZYNQ-7020 in a reasonable manner and studying the interconnection between various modules,design schemes for the main modules of the real-time target detection algorithm are proposed.Finally,based on the design schemes for the main modules,the real-time target detection algorithm is implemented on the ZYNQ platform,and the system is validated using real radar data.Firstly,the model of the main program is constructed,and the preliminary framework of the system is programmatically implemented.Then,program development for the PS side is completed,involving data reception,storing the data in the DDR3 of the PS,extracting a frame of data for transmission to the PL,and calculating the processing time required for a single frame of data.Next,program development for the PL side is completed,involving data arrangement,calibration,fast Fourier transform,DBF processing,and displaying the target detection results on an RGB display device.Finally,the system is debugged and experimentally verified.The real-time target detection algorithm designed on the ZYNQ platform meets the requirements of real-time target detection,and the test results demonstrate the rationality and effectiveness of the implemented algorithm.