资源描述：
《2013A Network Science Perspective of a Distributed Reputation Mechanism》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、Networking20131569702411123ANetworkSciencePerspectiveof45aDistributedReputationMechanism678RahimDelaviz,NielsZeilemaker,JohanA.Pouwelse,andDickH.J.Epema910DelftUniversityofTechnology,theNetherlands11Email:r.delavizaghbolagh@tudelft.nl121314Abstract—Reputationmechanismsarewidelyusedinonlin
2、edirectedgraphwithnodesrepresentingpeersandedgeweights15networkstorankusersorproducts,butdespitetheirimportance,representingamountsoftransferreddata.Thissubjectivegraph16veryfewstudieshavebeendoneorpublishedontheirrealisthenusedbyeachpeertocalculatethereputationvaluesofbehavior.Inthispape
3、r,westudyanInternet-deployeddistributed17otherpeersbyapplyingtheMaxflowalgorithmtothegraph,reputationmechanismcalledBarterCastthatisspecifically18designedforpeer-to-peerfile-sharingsystems.TheBarterCastinterpretingtheedgeweightsas“flows”.19mechanismisbasedonbuildingaweighteddirectedgraphfromT
4、ocollecttherequireddata,westartedaTriblernetwork20thedatatransfersthathaveoccurredamongthepeers,andoncrawlerinSeptember2010,whichstillisrunning;TriblerisaemployingtheMaxflowalgorithminthisgraphtoevaluaterepu-21BitTorrrent-basedpeer-to-peerfilesharingclientthatisusedtations.Inthispaper,westu
5、dythismechanismfromthenetwork22forpeer-to-peerfile-sharingandvideo-on-demandservices[2],perspectiveandweprovideadetailedanalysis,whichincludes23suchnetworktopologymeasuresasthedegreedistribution,nodeandthatusesBarterCasttorankpeers.Themaintaskofthe24interconnectivity,theclusteringcoefficien
6、t,communitystructure,crawleristodiscoverpeersandtocollectdatatransferrecords25anddistancemeasures.Besides,westudythegeographicalspreadfromthem.Usingthepermanentidentifiersofthepeersweareandcontentsharingbehaviorofthesystemparticipantsand26abletocorrectlygroupthecollectedrecordsfromdifferen
7、tcorrelatetheresultswiththeirconnectivityinthenetwork.We27peersandtogenerateaglobalnetwork,whichwecalltheinterpreteachevaluatedmeasureinthescopeofreputationand28file-sharingmechanismsandproposerelevantimplicationsandwork-graph.Moreover,intheTriblernetwork,therearefou
