资源描述：
《基于非合作动态博弈的网络安全主动防御技术研究_林旺群.pdf》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。
1、计算机研究与发展ISSN1000-1239PCN11-1777PTPJournalofComputerResearchandDevelopment48(2):306-316,2011基于非合作动态博弈的网络安全主动防御技术研究林旺群王慧刘家红邓镭李爱平吴泉源贾焰(国防科学技术大学计算机学院长沙410073)(linwangqun@nudt.edu.cn)ResearchonActiveDefenseTechnologyinNetworkSecurityBasedonNon-Cooperative
2、DynamicGameTheoryLinWangqun,WangHui,LiuJiahong,DengLei,LiAiping,WuQuanyuan,andJiaYan(CollegeofComputer,NationalUniversityofDefenseTechnology,Changsha410073)AbstractGametheory,animportantpartofartificialintelligenttechnique,hasbeenappliedonnetworkdefe
3、nseverywell.Staticmodelhasbeenusedwidelyinmostofthepreviousstudies.However,someworkshowssuchmodelcannotfollowtheevolvingofthestrategiesofattackers.Inthispaper,anactivedefensemodelbasedondynamicgametheoryofnon-cooperativeandcompleteinformationhasbeeng
4、iven,thatis,theattack-defensegametreehasbeengeneratedbyaddingsomevirtualnodesontheoriginalattack-defensegraph.Basedontheattack-defensegametree,thebestdefensestrategiesareachievedundercurrentnetworkenvironmentthroughresolvingtheNashequilibriumindiffer
5、entsituations.Besides,forthescenarioswithcompleteinformationandincompleteinformation,twoalgorithmshavebeenproposedrespectively.Theanalysisandexperimentalresultsshowthatthecomplexityofthealgorithmscanbeguaranteednotworsethanothersimilarworks.Moreover,
6、notonlyforscenariowithcompleteinformation,butalsoinincompletecases,thesensibleresultscanbefound.WiththecomparisonofmixedstrategyNashequilibriumgeneratedbystaticgameanddescribedinaprobabilisticform,resultsgivenbythesub-gameperfectNashequilibriumaremor
7、eeasilytobeunderstoodandoperated.Networkresearchbasedongametheoryshouldhaveagoodapplicationinthefuturenetworksecurityproduct.Keywordsnetworksecurity;activedefense;dynamicgametheory;attack-defensegametree;Nashequilibrium摘要目前基于博弈的网络安全主动防御技术大多采用静态博弈方式.针
8、对这种静态方式无法应对攻击者攻击意图和攻击策略动态变化的不足,基于非合作、非零和动态博弈理论提出了完全信息动态博弈主动防御模型.通过/虚拟节点0将网络攻防图转化为攻防博弈树,并给出了分别适应于完全信息和非完全信息两种场景的攻防博弈算法.理论分析和实验表明相关算法在复杂度不高于同类算法的前提下:1)不仅适应于完全信息博弈场景,而且在非完全信息的特殊场景下仍能够得到合理的解;2)与采用静态博弈给出的以概率形式描述的混合策略Nash均衡解相比,给出的从子博弈精炼Nash均衡中抽出的解具有更好的可理解性
