Millimeter Wave Radar Detection System Based On Fractional Fourier Transform

With the popularity of radar in civilian field,various target detection algorithms have emerged.Millimeter wave radar has been rapidly developed and applied to civilian radar because of its high resolution and low cost,which is able to relieve the effects caused by adverse weather.In traditional target detection algorithms,fast Fourier transformation(FFT)is the basis of many detection algorithms,such as moving target indicator(MTI)and moving target detection(MTD).For complex environment,detecting weak targets may occur false alarms or missed judgments due to low signal-to-noise ratio.Millimeter wave radar uses linear frequency modulation continuous wave(LFMCW).If fractional Fourier transform(FRFT)is used to process such non-stationary signals,the signal-to-noise ratio can be significantly improved,thus improving detect accuracy and reducing missed judgments and false alarms.In this paper,a millimeter-wave radar target detection algorithm based on FRFT is proposed,and the detailed flow and steps of the algorithm implementation are also given.In addition,the advantages and disadvantages of constant false-alarm rate(CFAR)are analyzed,and the improved greatest-of(GO)CFAR Radar signal algorithm are proposed.In order to apply the algorithm to the real project,the algorithm implements the hierarchical iterative search,which overcomes the problem of slow operation for the two-dimensional iterative method,and greatly improves the operation speed and better applies to the actual scene.In order to apply the multiple scenes,millimeter-wave radar(i.e.,IWR1642EVM)and data acquisition board(i.e.,DCA1000EVM)of Texas Instruments(TI)are selected by investigating the sensors of several companies,with which the whole system is built.Next,by combining the TI's software development kit(SDK)and Matlab,the debugging such as signal transmission,reception,storage,and signal processing requires to be completed,and the experimental data is collected in multiple scenarios.Finally,the experimental data collection operations,processes,and given results are presented in detail.Compared with MTD and FRFT-based MTD,we can achieve significative improvement of FRFT in the signal-to-noise ratio.The algorithm uses a fast algorithm of fractional Fourier transform.The calculation speed is slower than FFT,but it still be applied to the actual scene.