Research On Signal Processing In The FMCW Radar System

Frequency Modulated Continuous Wave Radar has been widely used in industry control,precise guidance,anti-collision systems,meteorological exploration,security systems and other related fields with the advantage of low power consumption,small blind detect area and high accuracy.In order to improve the performance of the FMCW systems,it is necessary to study and improve the signal processing.In this dissipation,based on the application of anti-collision systems of vehicles,the denoising,power spectrum estimation and multi-target detection have been analyzed.In the first part,the basic structure of FMCW Radar was described,analysis its working principle.The related parameters and indexes of the FMCW Radar and analysis each part of its modules was described.Besides,the measure principle of distance,velocity and angle and deduce the formula of them in the triangular linear frequency modulation continued wave Radar have been analyzed.In the parameters measurement of targets,the key points lay in the frequencies detection of targets and multi-targets detection.Therefore,in the section of frequencies detection,the method of denoising,power spectrum estimation and multi-target detection have been analyzed.In the section of denoising,the auto-correlation function to lower down the noise power and reserve most power of the useful signal was used.This method was analyzed theoretically and the denoising algorithm structure has been designed,which uses FFT to realize the fast computation of auto-realtion function.The simulation show that the method can lower the noise power effectively.In the section of power spectrum estimation,the classical and modern power spectrum estimation have been described.In order to improve the SNR,a low noise coherent accumulation structure based on the auto-relation function has been described.The structure can accumulate useful signal coherently and improve SNR.From the simulation shows that the structure can get 10log(M)dB SNR improvement,which M is times of accumulation.After the detection of frequency of targets,the ghost targets problem in the multi-target detection need to be solved.Firstly,the multi-target detection algorithm in traditional hybrid scheme(FMCW+CW)system has been analyzed and it has faults in some particular situations.Secondly,using the adjacent detection results and using the moving character of targets frequencies can improve the traditional method.Besides,a new velocity matching method was introduced,which uses the law of frequency moving.At last section,the FMCW Radar system was build up and some related DSP codes were write into the hardware board.Besides,using Matlab to build up a control and display platform.Based on these platforms,the experiments of measuring distance,velocity and angle parameter have been finished and it testifies the algorithm of low noise coherent accumulation power spectrum estimation and multi-target detection.