Design And Implementation Of Range And Angle Estimation Method Based On FMCW MIMO Radar

Millimeter wave technology has a wide range of applications in fields such as autonomous driving,aerospace,medical diagnosis,and radar target detection.By using multiple transmit and receive antennas,Multiple-Input Multiple-Output(MIMO)millimeter wave radar forms virtual elements in space to increase the aperture of the array,achieving high-resolution spatial target identification.Frequency Modulated Continuous Wave(FMCW)signal processing technology can effectively resist interference and noise,thereby improving target detection accuracy.Therefore,this paper applies FMCW technology to MIMO millimeter wave radar,and studies the design and hardware implementation of distance-angle estimation based on FMCW MIMO radar.The main work of this paper is as follows:Firstly,this paper analyzes the signal model and ranging principle of FMCW radar,and then derives the signal model and angle estimation principle of FMCW MIMO.The classic FMCW MIMO radar distance-angle estimation algorithm is studied,and the Direction of Arrival(DOA)algorithm is analyzed by simulation.Second,this paper analyzes the hardware design requirements for FMCW MIMO radar distance-angle estimation methods,proposes a method that first uses the twodimensional Fast Fourier Transform(2D-FFT)algorithm to roughly estimate the target’s distance-angle information,and then uses the single-parameter estimated Linear Chirp-Z Transform(CZT)algorithm and the Real-Valued Local Multiple Signal Classification(MUSIC)algorithm to accurately estimate the target range and angle.The proposed distance-angle estimation method is verified through simulation and real data processing,demonstrating its feasibility and effectiveness.The hardware implementation of the method is then designed,including selecting a development platform,choosing data formats,analyzing the implementation process,and dividing the design into modules.Finally,the proposed method is verified by processing real data using a Field Programmable Gate Array(FPGA)chip.The FPGA implementation of the singleparameter estimated CZT algorithm is the focus of the research,and the fixed frequency window refinement of frequency-unknown signals is achieved by using Fast Fourier Transform(FFT)and CORDIC algorithms to obtain accurate range values on the FPGA.The FPGA implementation of the Real-Valued Local MUSIC algorithm is also studied in detail,including using a linear transformation matrix to make the input data real-valued,calculating its real-valued covariance matrix,performing feature decomposition,and using the time-division multiplexing BLV structure and parallel single/double-sided Jocabi iterative algorithm to obtain its feature vector and eigenvalue,and inputting the feature vector and rough estimated angle into the local spectral peak search part to obtain accurate estimated angle value.In the simulation section,the proposed distance-angle estimation method is implemented on the FPGA,and the output results of the FPGA and simulation software are compared to verify the correctness of the method implemented on the FPGA.