Unmanned aerial vehicle individual multi-dimensional domain electromagnetic signal feature deep learning identification method

• Main Authors: ZHAO LANTIAN, QI PEIHAN, ZHOU XIAOYU, LI ZAN, ZHANG WEILIN, AHN JI-HYE, HE JINYANG, DING YUANLEI, ZHENG SHILIAN, JIANG TAO, WANG DANYANG
• Format: Patent
• Language: Chinese; English
• Published: 08.03.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses an unmanned aerial vehicle individual multi-dimensional domain electromagnetic signal feature deep learning identification method, and the method comprises the steps: obtaining an unmanned aerial vehicle electromagnetic signal sample through a radio monitoring technology, carrying out the labeling of the sample, and obtaining a signal sample set; performing multi-dimensional domain electromagnetic signal feature extraction on each signal in the signal sample set to obtain a multi-dimensional domain feature vector corresponding to each signal sample; performing multi-dimensional domain electromagnetic signal feature compression processing on the feature vector corresponding to each signal sample to obtain a feature vector after compression and dimension reduction; forming a training data set through the compressed and dimensionality-reduced feature vectors of all the signal samples; the deep learning classifier module constructs a classifier; training a classifier through the training d