{"title":"IPTV网络的u型转弯缓存优化","authors":"L. Sofman, B. Krogfoss, A. Agrawal","doi":"10.1109/SARNOF.2009.4850338","DOIUrl":null,"url":null,"abstract":"In an IPTV network, a large amount of unicast traffic from the Video Head Office (VHO) to subscribers requires significant BW and equipment resources in the network. Caching of video content allows reduction of bandwidth and IPTV network cost. In this paper, we consider a caching architecture in which part of the content is cached at the edge of the network (e.g., at DSLAMs), and the remaining part is stored in server(s) at the VHO. We assume that the community of subscribers subtended at each cache location (DSLAM) may have unique interests. Therefore, the list of cached items is customized for each cache location (e.g., individual DSLAM), and different cache locations should store different content. We consider U-turn caching where titles cached at a given DSLAM may serve the requests of subscribers that are subtended not only at this DSLAM but also from those subtended at neighboring DSLAMs. We discuss how the U-turn caching architecture allows increased cache effectiveness (hit rate) by sharing cache resources in the access nodes (DSLAMs).","PeriodicalId":230233,"journal":{"name":"2009 IEEE Sarnoff Symposium","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2009-03-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"U-turn caching optimization in IPTV networks\",\"authors\":\"L. Sofman, B. Krogfoss, A. Agrawal\",\"doi\":\"10.1109/SARNOF.2009.4850338\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In an IPTV network, a large amount of unicast traffic from the Video Head Office (VHO) to subscribers requires significant BW and equipment resources in the network. Caching of video content allows reduction of bandwidth and IPTV network cost. In this paper, we consider a caching architecture in which part of the content is cached at the edge of the network (e.g., at DSLAMs), and the remaining part is stored in server(s) at the VHO. We assume that the community of subscribers subtended at each cache location (DSLAM) may have unique interests. Therefore, the list of cached items is customized for each cache location (e.g., individual DSLAM), and different cache locations should store different content. We consider U-turn caching where titles cached at a given DSLAM may serve the requests of subscribers that are subtended not only at this DSLAM but also from those subtended at neighboring DSLAMs. We discuss how the U-turn caching architecture allows increased cache effectiveness (hit rate) by sharing cache resources in the access nodes (DSLAMs).\",\"PeriodicalId\":230233,\"journal\":{\"name\":\"2009 IEEE Sarnoff Symposium\",\"volume\":\"1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2009-03-30\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2009 IEEE Sarnoff Symposium\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/SARNOF.2009.4850338\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2009 IEEE Sarnoff Symposium","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SARNOF.2009.4850338","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
In an IPTV network, a large amount of unicast traffic from the Video Head Office (VHO) to subscribers requires significant BW and equipment resources in the network. Caching of video content allows reduction of bandwidth and IPTV network cost. In this paper, we consider a caching architecture in which part of the content is cached at the edge of the network (e.g., at DSLAMs), and the remaining part is stored in server(s) at the VHO. We assume that the community of subscribers subtended at each cache location (DSLAM) may have unique interests. Therefore, the list of cached items is customized for each cache location (e.g., individual DSLAM), and different cache locations should store different content. We consider U-turn caching where titles cached at a given DSLAM may serve the requests of subscribers that are subtended not only at this DSLAM but also from those subtended at neighboring DSLAMs. We discuss how the U-turn caching architecture allows increased cache effectiveness (hit rate) by sharing cache resources in the access nodes (DSLAMs).