Micro-motion Feature Extraction And Detection Of Wideband Radar UAV Target

Unmanned aerial vehicles(UAVs)are very common and widely used in military and civil fields,but there are risks of being misused for criminal acts and threatening people’s safety.In order to resist the threat of UAVs,it is very important to detect UAVs in time.This task is expected to be challenging as UAV have low Radar Cross Section(RCS)and fly at lower altitude and slower speed in comparison with conventional aircraft.The radar micro-Doppler signatures of a target depends on micro-motions such as rotations or movements of parts of the target other than the main body motion(e.g.,spinning rotor blades)and therefore characteristics can be obtained to help detect the target accurately.In order to detect the uav more accurately,this paper combines the traditional detection methods with the micro-Doppler feature extraction of the drone to achieve more accurate detection of the drone and reduce the probability of false alarm.In this paper,the concept and connotation of the micro-Doppler effect are introduced in detail.The micro-Doppler effect of the target plays an important role in the accurate detection of the target.The rotating rotor blades with more prominent micro-Doppler features were modeled and simulated,and the micro-Doppler effect was analyzed.Then,the time-frequency analysis method is used to extract the micro-motion features of the target,and the short-time Fourier transform,singular value decomposition and wavelet transform are used to extract the micro-Doppler features of the rotating rotor blades and the UAV target.Get their physical structure parameters.Next,in this paper,for the target detection problem of wide-band radar without distance extension,how to distinguish the target of the drone from the strong fixed ground clutter is more important.In this paper,the traditional CA-CFAR detection method and micro-Doppler feature extraction method are combined to propose an RD-spectrum detection method,which has better detection performance.Finally,in this paper,for the target detection problem of wide-band radar range extension,when the radar bandwidth is wide enough for the small drone to occupy multiple distance units,the micro-Doppler characteristics of the drone will be more obvious.In this case,we studied two extended target detection methods based on onedimensional distance image which are the M/N detection algorithm and the spatial scattering density generalized likelihood ratio(SSD-GLRT)detection algorithm.Then we introduced a range-extended target detection method in RD-spectrum,and compared these three methods.After that,the perturbation feature based on frequency modulation in the spectrum is extracted from the range-extended target,which can distinguish the target from clutter effectively.