Font Size: a A A

Research On The User Behavior Characteristic Based Storage Machanism For Interactive P2P VoD Systems

Posted on:2011-08-21Degree:MasterType:Thesis
Country:ChinaCandidate:J WangFull Text:PDF
GTID:2178360305476544Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the development of network technology and broadband network, the streaming media application has got lots of attention. The tranditional VoD systems are almost based on C/S (Client/Server) architecture. As the large growth of clients, the bandwith of the server will be the bottleneck of the system. The research and application show that the peer in P2P VoD systems can serve other peers with its bandwith. Since P2P allows sharing bandwith, storage and other resource among peers, it can significantly reduce the server load and greatly improve the utilization of network resource. On the other hand, as the heterogeneity of peers'service capabilities, the dynamic of peers and the specipality of streaming media, all these make the construction of P2P streaming media system face a lot of challenges.The interaction of P2P VoD systems needs the support of content storage, location and delivery. The existence of the available content is the premise of the implementation of interactive operation. So, how to reasonably use the cache and storage to distribute the content to peers is the key problem in the research. First, we make research on the users'behavior characteristic and a range constraint, popularity aware data prefetching scheme (RC-PAP) for peer-to-peer video is proposed according to the user behavior model of random seeking and the popularity of media data blocks. Then we introduce the popularity into caching mechanism and research the caching and replacement process on the data segments. Based on these, we design a selection and update algorithm on the basis of the utility (SDU). And then, according to the problem of unreasonable data distribution in the passive caching strategy, we discuss how to efficiently use the bandwidth and rescource of peers to replicate the needed data segments and then a replication strategy based on the incentive mechanism is proposed. Finally, we choose p2pstrmsim as the platform to do the simulation experiment and analyze the simulation result to verify the validity of algorithms proposed by this article.The results show that it is necessary to improve the interaction of the VoD system by cachinng and replicating content. It has important reference meaning for the P2P VoD system coming to mature.
Keywords/Search Tags:P2P, Video on Demand (VoD), Interaction, User Behavior Characteristic, Active Replication
PDF Full Text Request
Related items