Peer Rewiring In Semantic Overlay Networks Under Churn

Semantic overlay networks have been proposed as a way to organise peer-to-peer networks; peers that are semantically, thematically or socially similar are discovered and logically organised into groups. Efficient content retrieval is then performed by routing the query towards peer groups based on their likelihood to match the query. In this paper, we study the behaviour of semantic overlay networks that support full-fledged information retrieval in the presence of peer churn. We adopt a model for peer churn, and study the effect of network dynamics on peer organisation and retrieval performance. The overlay network is evaluated on a realistic peer-to-peer environment using real-world data and queries, and taking into account the dynamics of user-driven peer participation. Using this evaluation, we draw conclusions on the performance of the system in terms of clustering efficiency, communication load and retrieval accuracy in such a realistic setting.