資源描述：
《基于模糊-隱馬爾可夫模型的復(fù)合式攻擊預(yù)測(cè)方法研究》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、萬(wàn)方數(shù)據(jù)擊場(chǎng)景的識(shí)別和攻擊意圖序列的預(yù)測(cè)三項(xiàng)功能，將成為主動(dòng)防御的重要組成部分，并應(yīng)用于主動(dòng)防御實(shí)踐中。通過(guò)仿真實(shí)驗(yàn)可以看出：經(jīng)過(guò)訓(xùn)練的隱馬爾可夫模型比未經(jīng)過(guò)訓(xùn)練的隱馬爾可夫模型在對(duì)復(fù)合式攻擊行為的識(shí)別和預(yù)測(cè)效果更佳，而且在當(dāng)前四類主流的復(fù)合式攻擊預(yù)測(cè)方法中，本文提出的基于模糊一隱馬爾科夫模型的復(fù)合式攻擊預(yù)測(cè)方法在報(bào)警信息預(yù)處理、報(bào)警信息關(guān)聯(lián)等方面的綜合性能最優(yōu)。關(guān)鍵詞：報(bào)警信息關(guān)聯(lián)規(guī)則模糊評(píng)價(jià)法隱馬爾可夫模型復(fù)合式攻擊萬(wàn)方數(shù)據(jù)AbstractAsweentertheinformationage，therequirementsof
2、securityofinformationtransmission，informationstorageandinformationprocessingareofamuchhigherstandard．Thenetworksecuritynotonlyrelatestothesafetyofthecountry,thedevelopmentofeconomyandtheprogressofthescienceandtechnology,butalsorelatestothevitalinterestsofeveryone．Netw
3、orkisadouble-edgedsword；itnotonlyacceleratestheinformatizationofsociety,butalsobringsahugechallengefortheinformationsecurityissues．Inrecentyears，therateofnetworksecuritycrimesriseseveryyear．Particularly,withtheadventofonlinebanking，mobilebanking，e—commerceandotheronli
4、neservices，andthenetworksecurityissuescausedbytheconstructionofavarietyofprivatenetworks，thefollowingnetworksecurityproblemsbecomehotissuesaswell．Inthisstage，thetraditionalpassivedefenseCan’tadapttothedynamicchangesofnetworksecurity．Afterapplyingofthedefenseindepthoft
5、hemilitaryfieldtothesecurityareas，activedefenseemerges．Inthispaper,throughthestudyoftheexistedmulti—stepaRackspredictionmethods，theimprovedApriorialgorithmandfuzzyevaluationareappliedtothehiddenmarkovmodel，andtheapproachtoforecastingmulti-stepattackbasedonfuzzyhiddenm
6、arkovmodelisproposed．AfterthesemanticanalysisofrawalertsandfeaturesanalysispossessedbyaRacks，therawalertsarefusedintosuperalertaccordingtotherulesfirstly．TherealIntentionsofattackersarehiddenandCan’tbeobservedbyobserversdirectly．Buttherawalertscanbeobserveddirectly．In
7、thisregard，hiddenmarkovmodelisappliedtomulti‘stepattackspredictionmethods，inwhichthealertsareasobservationlayerandtherealintentionsareashiddenlayenThentheattackscenariowhichthealertsbelongtoisrecognizedbytheForwardalgorithmofHMMandthenextpossibleattacksequenceisforeca
8、stedbyViterbialgorithm．Finally,theexistedhiddenmarkovmodelistrainedbyBaum-WelchalgorithmofHMMandwegetanewfuzzyhiddenmarkovmo
