基于分布智能代理的自保护系统研究

《基于分布智能代理的自保护系统研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、一I虹面匝凹，⑩天串大蓦博士学位论文■■～一⋯一⋯■一一级学科：盐薹垫璺生兰塑茎举科专业：盐薹垫壁旦蕉茎作者姓名：—难亚垩——指导教师：啦训姐3P-Ⅻ+*_女岳＆ABSTRACTWiththedevelopmentofInformationtechnology,Networksecurityisgettingmoreandmoreimportant．Severalkeytechnologiesofnetworksecurityarestudiedandanalyzedsystematicallyinthisd

2、issertation．Theyarenetworksecuritymodel，themechanismofDistributedDenialofService(DDoS)Attackdefending，clusteringalgorithmsfornetworkintrusiondetection，informationencryptionalgorithm，andautonomicmtmsionresponse．Someinnovationshavebeenachievedandpresentedinth

3、isdissertation：Havingstudiedthecurrentlynetworksecurityandagenttechnology,thisdissertationpresentsaself-protectionmodelbasedonmulti-agentanddescribesthecomponentsandpropertyofthismodelindetail．DDoSiSoneofthegreatestmenacestoIntemet．Theexistedmechanismfordef

4、endingDDoScarl’tdistinguishnormalnetworkpacketsandattacknetworkpacketswhenthecontentsofnetworkpacketsofDDoSarerandomized．ThisdissertationpresentsamechanismfordefendingdistributeddenialofserviceattackwhichcandetectthepresenceofapotentialDDoSattackanddivertat

5、tacktrafficdestinedforthenetworkbeingmonitoredwithoutaffectingtheflowoflegitimatetraffic．EmulationshowsthatthismechanismfordefendingDDoSiSeffectiveandfeasible．Datamininghasuniqueadvantagesinacquiringunknownknowledge．So，intrusiondetectionbasedondataminingbec

6、omesahotissue．ThisdissertationpresentsanImprovedPartitioningAroundMedoids(IPAM)algorithmandevaluateditperformanceonthenetworkconnectiondatasets．ExperimentstudiesshowthatthisalgorithmiSfeasibleandeffectiveforunknownintmsiondetection．II'1thisdissertationaneww

7、ord—orientedstreamcipherWhichbasedonconventionalencryptionalgorithmispresented．Acompletedescriptionofthealgorithm，anevaluationofitssecurityproperties，performanceandimplementationaspectsaregiven．Thecryptanalysisofthisalgorithmdoesnotrevealallattackbetterthan

8、exhaustivekeysearch．TheSpeedofthisalgorithmisasfastascommonlyblockciphers．Atlast，acost-sensitivemodelbasedonreliabilitydegreeiSproposed．ThenallautomatedintrusionresDonsetechniquethatiSbasedOnmulti．sour