基于聚类等势星球图的物联网物理层认证方法

打开文本图片集

中图分类号：TN929.5 文献标志码：A文章编号：1003-3114（2025）04-0789-09

doi：10.3969/j. issn.1003-3114.2025.04.016

Abstract：Duetotheopennessofradio transmision，theinformationsecurityproblems broughtbywirelesscommunicationnetworksareconstantlyemerging.Inordrtoensurethesecurityofuserinformationinwirelesscommunicationanovelphysical-layer identityauthenticationmetodfortheInternetofThingshasbenproposed.ThismethodinitiallemploystheI/Qmodulationalgo rithmtotransfooe-dimensionalsignalsintotwo-dimensionalconstelationmapsonthecomplexplane.Subsequently，itcombines theDensity-BasedSpatialClusteringofAplicationswithoise（DBSCAN）algorithmtoconstructtwo-dimensionalolor-clusterd contourstellarimages.Theseimagesemphasizediferentchannelcharacteristicsandfacilitatesignalclasification.Basedontwo typesofmeasuredchnneldate：oneinvolvingdiferentphysicalregionswhereuserterminaldevicesmaybelongtodiferentcelular zones，andtheotherinvolvingmeasuedchaneldatawithinthesamecellularzone.Inbothscenarios，frequency-domainfull-channeldatawereconvertedintotwodimensioalimagesforrecognitiotesting.ConvolutionalNeuralNetwork（CNN）andSupportVec torMachie（SVM）modelswereemployedincomparativeexperimentson8OOclusteredcontourstelarimages.Simulationresults demonstrate that the proposed CNN model achieves an accuracy rate of 99.6% inuser recognition，effectively identifying legitimate users.Thisresearch providesasolidtheoreticalfoundationfortheapplicationofclusteredcontourstelarimagealgoritmsandCNN models in the field of physical-layer authentication.

Keywords：Internetof Things；information security；channel feature;clusteredcontour stellar images;CNN

0引言

随着移动通信技术的迅速发展，越来越多的通信设备加人无线通信网络，伴随着大量价值较高的重要信息在其中传输，通信安全问题成为贯穿整个无线通信技术发展过程中的首要问题，因此高效应对无线通信安全问题变得十分迫切，且具有重要的研究意义与应用价值[1]

物理层认证作为保障无线通信安全的核心技术之一，相比于应用层认证机制，能够有效抵御模仿攻击，具有兼容性好、复杂度低、认证速度快、不需要考虑各种协议执行的特点。（剩余9940字）