Study On Caching Algorithm Of P2P VoD Media Streaming

With the rapid development of the network technique, the media streaming technology has been used in network living broadcast ,VoD, remote instruction and so on. But the traditional C/S model brings the server on performance bottleneck and bandwidth bottlenecks,this makes it cannot be used in large-scale processes of operating concurrently. This years the new technology of P2P use the direct interaction between peer and peer to realize the sharing of information resources, greatly reduce the system's reliance on central servers. However, as VoD which demand a high degree of interactivity, makes achievement a higher degree of complexity. Now how to build a P2P media streaming VoD system with large-scale, high scalability, reliability and performance in a dynamic peer-to-peer network environment is a hot research topic in recent years.This dissertation proposes a new algorithm named GSP-K (Global Segment Popularity-Based Caching Algorithm - K) , which is a peer caching algorithm based on global segment popularity for P2P VoD system. This algorithm evaluate the cache utility of media segment via gathering the statistics of media segment popularity globally, and make the numbers of a media segment copies in the whole system varies directly with it's cache utility. The simulation experiments indicate that this algorithm has better performance in increasing statistics precision of segment popularity, cache space utilization rate and segment hit radio than others algorithms such as LRU, LFU and so on.