面向卫星通信的非线性选代学习混沌通信

打开文本图片集

关键词：卫星通信；混沌通信；选代学习；长短期记忆神经网络；混沌同步质量；参数优化中图分类号：TN911.7-34；TP391.41 文献标识码：A 文章编号：1004-373X（2025）10-0015-05

Abstract：Inalusion tothedemand forsecure transmission performanceinsatelitecommunication，especially in applicationscenarioswithoutsecurepayloads，achaoticcommunicationmodeloptimized bytheimprovednonlinear iterative learning isproposed.Themodeltrainingof theencryptedsignalmixedwithchaoticcariersignalsandrawinformationina specificratioisconductedbymeansof thelngshort-term memoryneuralnetwork（LSTM）toobtainneuralnetworkmodelhighly consistentwiththeparametersofthelasertransmiter，soastosolvetheproblemofincompletematching betweenthereceiver andtransmitersystemparametersinchaoticcommunicationsystems.Inordertofurtherimprovethesynchronizationqualityof chaoticsignals，iterativelearningisintroducedtoconducttheparameterotimizationoftheLSTM.Thedecryptionrecognition rateof theproposednonlinearchaoticcommunicationsynchronizationmodelbasedonimprovedLSTMisfinallstableat 94.07% ，which is 4.03% and 1.82% higher than those of the radial basis function （RBF） neural network chaotic secure communicationmodelandthechaoticsecurecommunicationmodelbasedonconvolutionalneuralntwork，respectivelyerifying that the proposed communication model has good comprehensive performance.

Keywords：satelitecommunication;chaoticcommunication；iterativelearning;longshort-termmemoryneuralnetwork; chaotic synchronization quality;parameter optimization

0 引言

近年来，全球各国航天技术发展迅猛，以星链和千帆为代表的低轨道通信卫星成为研究热点。（剩余5731字）