Research On Vehicle Millimeter Wave Radar Algorithm And Its Implementation In FPGA

Nowadays,with the popularity of cars,the frequent occurrence of traffic accidents has become a problem that people pay more and more attention to and hope to solve.The emergence and application of automotive radar can effectively avoid accidents.At the same time,with the rapid development of auto driving industry,the higher performances of the on-board sensors are required.Therefore,the research on vehicle borne millimeter wave radar has great potential value and practical significance.With the development of integration technology,FPGA not only has the advantages of miniaturization and low power consumption,but also its cost is decreasing.FPGA not only retains its traditional advantages such as diversity of communication interfaces and parallel computing,but also integrates a large number of soft and hard cores.It can complete radar signal processing independently.Therefore,the signal processing system of vehicle radar based on FPGA will become the future trend.Based on the basic working principle of sawtooth modulated FMCW millimeter wave radar,this paper studies two kinds of algorithms that can improve the performance of radar signal processing system.They are velocity expansion algorithm and high-precision DOA estimation algorithm.Velocity expansion algorithm includes velocity expansion algorithm based on Doppler phase offset compensation and velocity expansion algorithm based on Chinese remainder theorem.The first algorithm can double the maximum unambiguous speed of radar.But its disadvantage is that the radar must adopt TDM MIMO mode.Another algorithm is better than the former in the ability to expand the maximum unambiguous speed.But it needs to rely on multi frame data for speed expansion,which reduces the radar refresh rate.MUSIC algorithm and Capon algorithm are studied and their performance is analyzed.The two algorithms can effectively improve the angle resolution of radar.Their performance is affected not only by the radar aperture,but also by the snapshot number and SNR of the input signal.When the input signals are the same,the performance of MUSIC algorithm is better than that of CAPON algorithm.However,MUSIC algorithm has higher complexity.And its performance is affected by the accuracy of source number estimation.In addition,the spatial smoothing algorithm and CFAR algorithm are also studied.Some of the above algorithms are implemented in AWR1642.According to the working principle of FMCW radar and the final performance of radar,this paper designs the hardware platform and signal processing system of vehicle borne radar which can be used for range,velocity and angle measurement.This paper determines the parameters of each link of signal processing.The simulation of the whole signal processing system in MATLAB verifies the feasibility of the system.At the same time,an echo signal model which can customize the target parameters is built by MATLAB.This data model is also used to verify the accuracy of the signal processing system in FPGA.Finally,based on xc7k325tffg676-2 FPGA,the signal processing system of vehicle radar with multi-objective ranging,speed measurement and angle measurement functions is realized.Next,we need to design the control module of front-end RF chip and the output module of measurement results in FPGA.We can also try to implement extended algorithm in FPGA to improve radar performance.