A New Approach for Achieving Traffic-Exchange Localization in P2P-based Content Distribution

摘要

Due to the fact that P2P applications have dominantly accounted for the entire Internet traffic, how to efficiently manage P2P traffic has become increasingly important. It has been recently proposed that the underlying network information can be shared between ISPs and P2P service providers in order to achieve efficient resource utilization, with the locality-based peer selection being a specific example. Based on such collaboration, we propose a proportional traffic-exchange localization scheme for making efficient use of network resources. Our approach employs locality information in order to regulate the volume of traffic exchange between peers according to their physical distance between peers. The key objective of our approach is to further reduce both intra- and inter-autonomous system (AS) traffic compared with basic locality-based peer selection solutions. Our simulation-based results have shown that this approach is not only able to reduce a significant of inter-AS P2P traffic, but also to balance the network utilization in comparison to existing approaches.