Research On Detection And Tracking Of UAV Using LFMCW Radar

Linear frequency modulation continuous wave(LFMCW)radar with the advantage of all-weather,high precision and no blind spots is an effective sensor to realize the detection of near-range unmanned aerial vehicle(UAV).However,the current detection of "low-small-slow" UAV is still facing a series of problems,for example,the clutter intensity of the ground architectural is generally much larger than the UAV,making the detection difficult.The flight speed of UAV is so slow that its Doppler spectrum and clutter spectrum are difficult to distinguish,which makes the speed measurement difficult.The arbitrary flight path of UAV with various modes such as emergency stop,rapid flight,ascending,descending and turning makes it is difficult to be tracked especially in the background of dense clutter environment.How does the LFMCW radar measure the UAV's performance? How to achieve strong clutter cancellation? How to achieve effective tracking of multiple UAVs under the condition of dense clutter? These are all issues that need to be resolved urgently.In this paper,the LFMCW radar is systematically studied for its detection and tracking technology of the UAV.Firstly,the two-dimensional Fast Fourier transform(2D-FFT)based on range and velocity is proposed,the time and frequency domains analysis of the beat frequency signal of LFMCW radar is investigated,and the key factors affecting the performance of the target speed measurement are studied.The radar signal characteristics and radar detection performance of the UAV target under clutter background(or not)are theoretically analyzed and verified by simulation.This paper focuses on the investigate the detection performance of low-small-slow unmanned aerial vehicle(LSS-UAV)in the presence of dense architectural clutter,and the key factors affecting the performance of the target speed measurement are studied.By using Monte-Carlo simulation,the influence mechanism of radar parameters(frequency modulation ratio,bandwidth,moving target indication or not etc.),target parameters(RCS,velocity etc.),and clutter parameters(amplitude,spectral width etc.)on detection performance are analyzed,the interaction mechanism between the three types of the parameters is also analyzed.The study shows that the detection in velocity dimension is crucial to discriminate LSS-Target.Meanwhile,it is better to use large bandwidth signal and long-time accumulation technology to improve the target detection performance on strong clutter background.Secondly,based on the piecewise time-frequency analysis and Hough transform,a new UAV detection algorithm based on time-frequency synthesis is proposed,including segmented time-frequency analysis,spectrum splicing,spectrum superposition and fixed target cancellation,Hough detection and other methods.The simulation results show that the time-frequency image splicing and superposition can be very convenient to realize the multi-target detection efficiency,while the time-frequency diagram is subtracted,which can realize the function of moving target indication(MTI)at the time-frequency level,which is beneficial to achieve fixed clutter suppression.Finally,based on the measured data by LFMCW radar,the signal processing of the measured data of LFMCW radar data,and the data processing analysis of the LFMCW radar based on multi-parameter synthesis are carried out.Many tracking countermeasures should be taken for practical factors such as multi-target/strong clutter/radar sidelobe effect/target maneuver etc.Specifically.In the initial stage of the track,a method of three-frame initial trajectories controlled by parameters is used to suppress a mass of side lobes and ground clutter caused by the range close of the radar.In the correlation stage,the associated gate and continuous prediction are adjusted to adapt to the low accuracy and missing points of the UAV measurement.In the filtering process,interactive multi-model(IMM)algorithm is adopted to adapt the random maneuver of the UAV.