Optimal decomposition of P2P networks based on file exchange patterns for multimedia content search replication

摘要

In this paper, we address the issues of multimedia content search and replication over peer-to peer networks. We use the concept of semantic proximity that exploits the file exchange patterns exhibited among peer users so as to decompose the network into semantic clusters. Peer nodes are then decomposed into semantic clusters so that a) the probability that a node locates content within its own cluster is maximized, while simultaneously b) the respective probability of finding this content outside this cluster is minimized The semantic organization is then used for applying efficient cluster-based content replication strategies. Two different schemes are examined; the unrestricted and the restricted approach. The first one distributes multimedia content within peers of a cluster so that the average hop distance among nodes and objects weighted by the object popularity is minimized. This approach takes into account no Quality of Service (QoS) guarantees. Instead, the second scheme iteratively partitions the nodes of a cluster during each object replication based on a QoS violation criterion. The proposed algorithms are experimentally evaluated and compared with other approaches to demonstrate the efficacy of the proposed schemes. Copyright 2007 ACM.