面向深度残差网络的移动目标防御决策算法研究

打开文本图片集

中图分类号：TN918.1 文献标识码：A文章编号：1009-3583（2025）-0098-05

Research on Mobile Target Defense Decision Algorithm Based on Deep Residual Network

CAI Jie-feng

（School of Information Engineering，Minnan University of Science and Technology， Quanzhou 36220o， China）

Abstract：Inthurretnewoksecuritycndefeeproeseeishenmatitisastttckddilttdefend Inordertorealizemovingtargetdefenseandenhancenetworksecurityamovingtargetdefensedecision-makingalgorithmontedto deepresidualetworkisroposed，icharks tecaractesticampledataschasobilargettaffchracteristicndetwork systemlogsateiputvectorofthedepresidualnetworkndbtains teoutputvectorafterthtreeresidualfullyonectedlaes ofthedepresidualnetworkrainandleanthinputvectorAccording toteobtainedoutputvector，theoutputlaersolvesthemoving targetbehaviorprobabilityvalueofthe movingtarget trainingsampledata.Underthisprobabilityvaue，theSoftmax layerealizeste clasificationofanoalmoving argetsandnoalmovingtargetsAccordingtothejectivefunctionsofboththeoffnsiveandde fensivesidesinthetime game，theoptimalmulti-stageMarkowtime gameresultissolved.Theexperimentalresultsshowthathealgorithmcanefectivelyidentifyteseuritystateofthemovingtargetandobtainthedefensedecisionofthemoving argetwiththegreatest benefit from the game.

Keywords：depresidualnetwork;mobiletargetdefense;decisionalgorith;timeae;Mrkowdecision;ofensiveandefensiveconfrontation

随着网络的普及和发展，互联网信息系统网络安全成为现今社会信息系统发展需要攻克的难题[1-3]。（剩余6930字）