欢迎来到天天文库
浏览记录
ID:37337588
大小:10.99 MB
页数:103页
时间:2019-05-22
《互联网内容分发优化问题研究》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、华中科技大学博士学位论文互联网内容分发优化问题研究姓名:田臣申请学位级别:博士专业:信息与通信工程指导教师:刘文予20081102华中科技大学博士学位论文量不必要流量。针对这个问题,本文考虑到P2P业务存在本地相关性,引入节点的地域信息和CDN网络对于P2P网络的协作能力,提出基于接入提供商的增强型邻居选择算法和基于CDN协作协议的邻居选择算法。实验结果显示,这两个算法都可以极大的减小接入提供商核心网流量和接入提供商间的流量。本论文所作的研究将对内容分发网络的进一步发展起到积极作用。关键词:内容分发;
2、对等网络;内容分发网络;数学规划;整数规划II华中科技大学博士学位论文AbstractWiththerapidincreaseofmultimediaservicesandInternetusers,InternetContentProvidershavetofaceenormousrequestsfromgiganticusergroups.ItishardfortraditionalClient/Serverbasedinfrastructuretomeetthesechallenges:serv
3、ers’powerandbandwidthoftheiraccesspointsarealwaysthebottlenecksofsystemscalabilityandperformance.Thissituationcallsfornewtechnologies.Twodedicatedcontentdistributionnetworksemergedtotaketheroleofcontentser-vicing.Oneispeer-to-peernetworks:usersintheInte
4、rnetformanoverlaynetworkandcooperativelydistribute/servicethecontents.ThemostimportantcharacteristicofP2Pisdirectresourcessharingbetweenusers,hencesolvethebottlenecksofservers’scalabilityandbandwidthlimitations.Thekeytechnologyislocationalgorithmfordist
5、ributedobjects.AnotherisContentDeliveryNetworks,whichisdeployed/maintainedbycommercialoper-ators.CDNnetworkspushcontentstotheedgeofthenetworks,hencereducecorenetworktrafficatthesametimereduceuserperceivedresponsedelay.Therearemanychallengesintheresearchof
6、contentdistributionnetworks.Andtherearestillspacesforfurtherimprove-mentformanyalgorithms,asoptimizationtheoryandtechnologyarenotwellexploitedyet.Theworksofthisdissertationconcentrateonoptimizeddesignofsomeaspectsofcon-tentdistributionnetworks,including
7、:optimizedresourcequeryalgorithmdesigninunstruc-turedP2Pnetworks;onlinereplicaplacementschedulinginCDNnetworks;cooperationbetweenP2PandCDNnetworks.Thecontributionsofthedissertationare:(1)Designdefectsofpreviouscontrolled-floodingalgorithmsareanalyzedfirst
8、;basedonoptimizationprincipal,anovelSelectiveDynamicQueryalgorithmisproposed.ThisalgorithmcanachievethesametrafficcontrollevelaspreviousDynamicQueryalgorithm,togetherwithapproximateresponsedelayasExpandingRing.Byextensiveexperiment
此文档下载收益归作者所有