{"title":"Individual Recognition of Communication Emitter Based on Deep Learning","authors":"Jie Xu, Weiguo Shen, Wei Wang","doi":"10.1109/ISAPE.2018.8634244","DOIUrl":null,"url":null,"abstract":"In view of the individual recognition problem of the communication emitter, this paper, starting with the subtle characteristics of the communication emitter in the signal layer, proposes a method of individual recognition based on deep learning. First, a recognition framework based on deep learning is established, and a convolution neural network containing two hidden layers is designed to extract local features through two layers convolution operations. Secondly, the stochastic gradient descent method is used to optimize the parameters, and the soft max model is used to determine the output label. Finally, the effectiveness of the method is verified by experiments.","PeriodicalId":297368,"journal":{"name":"2018 12th International Symposium on Antennas, Propagation and EM Theory (ISAPE)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 12th International Symposium on Antennas, Propagation and EM Theory (ISAPE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISAPE.2018.8634244","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
In view of the individual recognition problem of the communication emitter, this paper, starting with the subtle characteristics of the communication emitter in the signal layer, proposes a method of individual recognition based on deep learning. First, a recognition framework based on deep learning is established, and a convolution neural network containing two hidden layers is designed to extract local features through two layers convolution operations. Secondly, the stochastic gradient descent method is used to optimize the parameters, and the soft max model is used to determine the output label. Finally, the effectiveness of the method is verified by experiments.