基于多域特征融合的雷达辐射源个体识别方法

打开文本图片集

中图分类号：TN957.51-34 文献标识码：A 文章编号：1004-373X（2025）17-0021-08

DOI：10.16652/j.issn.1004-373x.2025.17.004引用格式：，，，等.基于多域特征融合的雷达辐射源个体识别方法[J].现代电子技术，2025，48（17）：21-28.

Abstract：Thecommon radar emiter individual identification methods based onthe time-frequencydomainof the signal failtofullycharacterizetheindividualcharacteristicsofradar，andtheyarefarawayfrommetingthenedsofradarndividual identificationinthecomplexelectromagneticsignalenvironmentsanddiferentradarworkingmodes.Inviewoftheabove，this paperanalyzesandcalculatesthecharacteristicsofthsignalontheime-domain，frequency-domain，andtransformatiodomain onthebasisoftetime-frequencydistributioncharacteristicsofthesigal，andtencomprehensivelyconstructsaulti-oain featuredatabaseofradaremiters.Amethod foroptimizing themaximumrelevanceandminimumredundant （MRMR）multidomainfeaturesisproposed.Adual-inputneuralnetworkrecognitionmodelbasedonmulti-dmainfeaturefusionisdesigned inovatively，andtheatentionmechanismisusedtoadaptivelypayatentiontotheweightsofkeyadarfeaturesindiferent signalenvironments，soastoeliminatethelowrecognitionrateoftime-frequencydistributionwhenonlythetime-frequency domainisused.Aftertheanalysisandcomparisonofexperimentsonrealtimereconnaissane，itisverifiedthattheproposed methodcanadapttothesimultaneousmulti-signalapplicationscenariossatisfactorily，andtheaccuracyrateof individual recognition isimproved bymore than 12%

Keywords：subtlefeatureofsignal;time-frequencydistribution;MRMR;atentionmechanism;dual-inputneuralnetwork; individual recognition

0 引言

雷达辐射源识别又称为特定辐射源识别（SpecificEmitterIdentification，SEI），可以准确地提供敌方雷达的重要军事信息，旨在唯一识别所有相同平台场景下的雷达辐射源个体，为军事行动中的电子保障措施系统提供了关键情报，具有广阔的应用前景。（剩余7725字）