Ryan LaFortune, C. Carothers, W. D. Smith, J. Czechowski, Xi Wang
{"title":"Simulating Large-Scale P2P Assisted Video Streaming","authors":"Ryan LaFortune, C. Carothers, W. D. Smith, J. Czechowski, Xi Wang","doi":"10.1109/HICSS.2009.908","DOIUrl":null,"url":null,"abstract":"In this paper, we show through detailed simulations that BitTorrent can be used to assist a server or content distribution network (CDN) for large-scale streaming delivery. Specifically, a content owner can distribute a file to over 131,072 users/peers with a peak swarm size of 16,384 concurrent users and a CDN bandwidth savings of 73%, while achieving an average user buffer time of under 2 seconds, which is an A+ on the StreamQ user performance rating system. Current research and even deployed systems rarely have swarms in excess of 1,000 concurrent peers for a single piece of content. Consequently, this is the first study that reports on peer-to-peer streaming for swarms of this scale. Further, the simulation model demonstrates that a high quality of service can be provided, while significantly reducing the distributor's transit costs for these large swarms.","PeriodicalId":211759,"journal":{"name":"2009 42nd Hawaii International Conference on System Sciences","volume":"171 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2009-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2009 42nd Hawaii International Conference on System Sciences","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/HICSS.2009.908","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
In this paper, we show through detailed simulations that BitTorrent can be used to assist a server or content distribution network (CDN) for large-scale streaming delivery. Specifically, a content owner can distribute a file to over 131,072 users/peers with a peak swarm size of 16,384 concurrent users and a CDN bandwidth savings of 73%, while achieving an average user buffer time of under 2 seconds, which is an A+ on the StreamQ user performance rating system. Current research and even deployed systems rarely have swarms in excess of 1,000 concurrent peers for a single piece of content. Consequently, this is the first study that reports on peer-to-peer streaming for swarms of this scale. Further, the simulation model demonstrates that a high quality of service can be provided, while significantly reducing the distributor's transit costs for these large swarms.