defending against malicious nodes using an sym based reputation system

defending against malicious nodes using an sym based reputation system

ID:7266805

大小:620.27 KB

页数:7页

时间:2018-02-09

defending against malicious nodes using an sym based reputation system_第1页
defending against malicious nodes using an sym based reputation system_第2页
defending against malicious nodes using an sym based reputation system_第3页
defending against malicious nodes using an sym based reputation system_第4页
defending against malicious nodes using an sym based reputation system_第5页
资源描述:

《defending against malicious nodes using an sym based reputation system》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库

1、DEFENDINGAGAINSTMALICIOUSNODESUSINGANSVMBASEDREPUTATIONSYSTEMRehanAkbani,TurgayKorkmaz,andG.V.S.Raju{rakbani@cs.utsa.edu,korkmaz@cs.utsa.edu,andgvs.raju@utsa.edu}UniversityofTexasatSanAntonioSanAntonio,TX,USAABSTRACTfuturebehaviorofanodebyanalyzingitspastbehavior.Ea

2、chnodethatinteractswithanothernodestoressomeManynetworks,suchasP2Pnetworks,MANETs,filefeedbackabouttheinteraction.Theinteractioniseithersharingnetworks,andonlineauctionnetworksrelyonclassifiedaslegitimateorsuspicious(orsomevalueinnodecooperation.Ifamaliciousnodegain

3、saccesstosuchbetween).If,forinstance,theinteractionconsistedofanetworkitcaneasilylaunchattacks,suchasspreadingdownloadingafile,theclientcoulddetermineifthevirusesorspam,orattackingknownvulnerabilities.downloadedfilewasindeedtheonerequested,oritwasaReputationSystems(

4、RS)defendagainstmaliciousnodesTrojan,avirus,orspam.Basedonthefeedbackofvariousbyobservingtheirpastbehaviorinordertopredicttheirnodes,anewnodecandecidewhethertotransactwithafuturebehavior.TheseRSsusuallycompriseofstatisticalgivennodeornot,eventhoughtheymayneverhavemo

5、delsorequationsthataredesignedbyhandandonlyinteractedbefore.eBayutilizesthisformofreputationdefendagainstspecificpatternsofattacks.Inthispaper,systemwhereusersleavefeedbackaboutotherusers.weproposeaSupportVectorMachines(SVM)basedRSthatdefendsagainstmanypatternsofatt

6、acks.ItcanbeIngeneral,wecansummarizeexistingRSs[1-3,5]withinretrainedtodetectnewattackpatternsaswell.WediscussageneralframeworkdepictedinFig1.AccordingtothisthechallengesassociatedwithbuildingRSsandhowourframework,anodethatneedstodecidewhethertotransactRStackleseach

7、ofthem.WecomparetheperformanceofwithanothernodeornotmustfirstgatherhistoricaldataourRSwithanotherRSfoundintheliterature,calledaboutthatnode(e.g.,theproportionofgoodvs.badTrustGuard,andperformdetailedevaluationsagainstatransactionsinthelastxminutes).Thenitappliesavar

8、ietyofattacks.TheresultsshowthatourRScustomizedmathematicalequationorstatisticalmodeltosignificantlyoutperformsTrustGuard.Evenwhenthetheda

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。