Research On Topology Optimization Of Bittorrent-Like Peer-To-Peer Systems

With the rapid development of Internet, lots of new network application systems appear. As an efficient content localization and distribution tool, Peer-to-Peer (P2P) file sharing systems have occupied a majority of Internet traffics. Whereas content localization has raised a lot of attention in the last few years, content replication has started to be the subject of active research recently. BitTorrent is a very popular and scalable file distribution system based on P2P paradigm. It has been successful at distributing large files quickly and efficiently without overwhelming the capacity of the origin server.Peer-to-peer (P2P) applications have shown their popularity in the Internet, while become a burden of Internet Service Providers (ISPs) because they produce huge traffic on network. The enormous traffic comes from the fact that when P2P systems construct overlay networks, they ignore the underlying network topology, so most of the connections are between remote peers and cross different networks. Take BitTorrent for instance, it builds its overlays by randomly selecting peers, a fact that has the potential to seriously handicap both individual performance and waste network resources. Study on BitTorrent[1] shows the fact that, for data already existing on local networks, parts of 70% to 90% are still downloaded from external networks. Such ignorance of network localization and randomly neighbor selection scheme, lead to huge cross-network traffic, which consume valuable ISP bandwidth.Solutions to this problem at present are less practical, but may easy to realize between IMS Terminals when IMS are implemented ubiquitously. We achieve this by constructing the NIDA (Network ID Aware) approach in a BitTorrent-like P2P application. Simulation experiments demonstrate that NIDA can decrease cross-network traffic effectively, not to mention it is easy-realized.