Load-Balancing Content Distribution in Structured Peer-to-Peer Networks

Multimedia streaming distribution over peer-to-peer networks stresses both the server and network because of large volume of data, high bandwidth requirement, and many concurrent peers. While the server limitation can be circumvented by replicating the data at more peers, the network resource limitation is far less easy to cope with, due to the difficulty in determining the cause and location of congestion and in provisioning extra resources. In this paper, we present a novel scheme for streaming distribution that partitions the client-peers into appropriate groups, so that the congestion-level of each session are reduced and also well balanced. The core problem is stated as partitioning the clients into disjoint subsets according to the degree of interference criterion, which reflects network resource usage and the congestion among the concurrent connections. Through simulation, we show that the algorithm is simple yet effective in achieving the design goals, particularly, (1) congestion balancing in the network, (2) the network bandwidth usage.