Community Detection Research And Application Based On P2P Log

With the high-speed development of the Internet and new Internet media. This explosive growth has not only led to the high cost to the bandwidth of the resource providers, but also requires network equipment providers to upgrade the network and improve the network transmission speed. P2 P network is a kind of software architecture which break the traditional client-server model, forming a client-client(as the server) model, and this form can improve the bandwidth resource utilization and ease the pressure on resources service providers.Community detection technology is based on complex network. In real life, many systems form a wide variety of networks, such as social networks, P2 P network and so on.According to the observation and research of these network, and then puts forward the concept of community detection.Community detection refers to a more reasonable and effective way to convert the chaotic network to reasonable community structure which takes the nodes and the edges between nodes as input, the network characteristics or some algorithm as a way. This can help people to find or improve the interatciont of community.The development of P2 P network technology is more and more mature, but the application is not good in practice. Firstly, it is unstable and is interrupted in the request resources process.Secondly, less effectiveness, this is because the bandwidth and processor hardware such as processing ability limit. Through combining community detection with P2 P technology, we can research a suitable detection algorithm for P2 P network, and find out the key information of the P2 P network which can improve the stability and effectiveness.In this paper, through the application of P2 P video-on-demand system and collection of P2 P log. We can research community detection methods, analysis P2 P log and use community detection methods to find out the key information of P2 P network. So form more reasonable P2 P network and simultaneously video-on-demand system also demonstrate the effectiveness of community detection methods.