Investigations On The Signal Processing Techniques For 77GHz Automotive Anti-Collision Radars

Recent years,automatic driving technology has attracted more and more attention.Millimeter wave car radar is one of the key technologies of automatic driving,it has the advantages of stable performance,high measuring accuracy and low cost.In this paper,the research on the signal processing part of the anti-collision radar system of automobile is carried out,including the construction of the signal processing platform and the research of the related distance measurement.The main work of the paper is as follows:Firstly,the background and current situation of the anti-collision radar are summarized,and the basic principle and overall scheme of the 77GHz anti-collision radar are introduced.The communication design between DSP board and millimeter wave front end based on OMAPL138 platform is studied,and the communication between OMAPL138 and HMC703,AD8285,AD8334 and gyroscope is studied.This paper introduces the distance measurement,velocity measurement and angle measurement of FMCW radar,and gives the simulation of MATLAB and Code Composer Studio.The final algorithm implementation can be run correctly on the OMAPL138,which provides high accuracy.Finally,this paper introduces the optimization of radar ranging algorithm.Due to the fence effect,the FFT can not provide accurate signal frequency.This paper introduces several commonly used frequency estimation algorithms and and their frequency offset estimation errors.On this basis,a hybrid frequency estimation algorithm based on FFT coefficients and M-Rife is proposed and implemented.The hybrid algorithm uses the FFT coefficients to avoid the problem that the individual amplitude information is susceptible to noise interference.In the calculation of frequency offset,the use of M-Rife algorithm to ensure the accuracy of the frequency estimation.