| The term ‘micro-motion' is defined as the vibration,rotation,precession with small amplitude and other high-order motion except the bulk translation of radar targets.The micro-motion of a target can well reflect its sophisticated features and the details of its motion,and provide new approaches for targets classification and recognition.Vibrating and rotating targets are mainly researched in this article.The formulas of radar echoes are deduced.The influence of the parameters for radar and targets on the Doppler frequency is investigated.For the Synthetic Aperture Radar,the effects of micro-motion on imaging are analyzed and the conclusion that micro-motion may result in the distance offset in range and the paired echoes in azimuth is obtained.Then,several algorithms are studied,which are applied to extract micro-Doppler features and estimate parameters of the single point targets,multiple point targets,vibrating non-rigid targets and helicopter rotors,respectively.The major contributions of this article are as follows:(1)For single point targets,a micro-Doppler characteristics extraction and parameters estimation algorithm based on phase-analysis method is proposed.It is a more efficient and convenient approach to estimate the micro-motion amplitudes,frequencies and initial phases of targets,whose micro-motion displacement is a single-frequency or multi-frequency signal.Moreover,it can be applied in the case where the airborne radar moves along the desirable or undesirable trajectory.Simulation results show its good performance on anti-noise.Compared with peak detection method and DFRFT-based method,the phase-analysis method provides higher estimation precision,a larger range of estimable micro-motion frequency and lower computation complexity.(2)For multiple point targets,an algorithm based on inverse Radon transform is studied.It is capable of separating and extracting micro-Doppler characteristics of multiple targets,even in the low signal-to-noise ratio circumstances.Compared with CEMD and Viterbi algorithm,this method can even extract the Doppler frequency and estimate its parameters of each component from the time-frequency diagram with multiple Doppler frequencies seriously overlapping.(3)For vibrating non-rigid targets,the three-dimensional geometric model built by FEKO and the point-array model are used to simulate the vibrating non-rigid target and its Synthetic Aperture Radar returns.The modulation effects of vibration on the radar echo of non-rigid targets are researched.Based on it,the autocorrelation function method and the peak detection method are applied to estimate the vibration frequency of the non-rigid targets.Simulation results indicate that the peak detection method is more suitable for non-rigid targets.(4)For rotating helicopter rotors,an approach to estimate the number,length and rotating frequency of blades is provided firstly.Considering the condition where radar scans multi-directionally so that targets may be exposed to radar beam only for a moment during each radar scanning period,an effective method to estimate parameters of helicopter rotor in short dwell time is proposed.By applying this method,the rotation frequency can be estimated accurately from a time-frequency diagram with the Doppler frequencies lasting for less than one rotation period.Compared with the traditional Viterbi algorithm and the peak detection method,the proposed modified Viterbi algorithm provides a more acceptable numerical complexity and estimation accuracy.In summary,the micro-Doppler characteristics extraction algorithms which are applied for targets in different scenarios are distinct.This paper focuses on the analysis of the micro-Doppler modulation of radar echoes for micro-motion targets in different scenarios,aiming to figure out the corresponding micro-Doppler characteristic extraction and parameter estimation algorithms,which provides a theoretical basis for the subsequent work about target classification,recognition and imaging. |
PDF Full Text Request