欢迎来到天天文库
浏览记录
ID:36565376
大小:2.50 MB
页数:60页
时间:2019-05-12
《基于融合思想的改进IDS的贝叶斯网络解法》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、东北大学硕士学位论文基于融合思想的改进IDS的贝叶斯网络解法姓名:何光宇申请学位级别:硕士专业:计算机应用技术指导教师:赵海20051201东北大学硕士学位论文AbstractBayesNetworkSolutiontotheImprovedIDSBasedonInformationFusionAbstractFacinglotsofchallengeandthreaten,intrusiondetectionsystemhasbecometheimportantpartofnetworksecuritydefendsystemwi廿1theactivedefendability.Ho
2、wever,thequickdevelopmentofnetworkandcomputertechnologies,bringtheunheard-ofchallengetothe扛aditionalIDS.Theincreasinglybrilliantandcomplexintrusionmeasures.makethetmditionalIDSincapabletofacetheintricateattackssuchasthemulti—stepsattacks.Theisolateddetectionandanalysismakealotoffalsereportandthe
3、failingreport,influencetheutilityseriously.ContraposingthedeficiencyofthetraditionalIDS,associatingwiththedevelopmenttrendoftheIDS,thisthesisbringoutamodelofimprovedIDSbasedoninformationfusion.Thismodelusethebasicprintipalofinformationfusion,dividetheIDSintomulti—layers.Thedamgatheringlayergathe
4、rinformationusingmultisensors.Pixellevelfusionlayerverifyandassembletherawdada,thentheFeaturelevelfusionlayermaketheprimarydetectionandassociationtoimprovetheabilityofrecognizingthecomplexintrusion.ThemajorworkinthisthesisareintheDecisionlevelfusionlayer,trytoratiocinateandanalyzetheintrusioninf
5、ormation.First,usethetrainingdataandpriorknowledgetolearnthebayesnetworkstructure,feudtheconsequenceofthefactorswhichinfluencethenetworksecurity.Second,usethetrainingdatatolearntheCPT,findtheprobabilityrelations.Thenusethedatafromthelowerlayerasthetestingdatatomakedecision.Givetheresultastheform
6、atofprobabilitytodecreasethefalsereportrateandthefailingreportrate,providethebeRersupporttotheadministrator.Inordertomakethemodeltofollowtheconductofthehost,makeon-linestudyofthebayesnetworkstructurewhilemakinganalyingusingthetestingdata.Besides,usingthefeedbackinformationfromtheexperttomodifyth
7、elearningconductiontofittheneedofthesystem.Finally,therespondlayertakedifferentmaneuvertothedifferenttypeandIevel.Keywords:Intrusiondetection;Informationfusion;Bayesnetwork;Statetransition;ProtocolanalysisIll独创声明本人声明所呈交的学位论文
此文档下载收益归作者所有