PeerView: query processing based on views over collaborative peers

Peer-to-peer computing is gaining more and more significance in both academia and industry, due to its widespread use currently and potential deployments in future applications. In this paper, we propose a prototype P2P application providing a view-based query processing mechanism in P2P environments, which is named as PeerView. A large number of peers may work under a collaborative and self-organizing scheme to answer their common queries or sub-queries. Given the workload and storage of each peer, some certain peers establish and maintain the views via a cost-based and negotiating way. When processing queries, the peers can first retrieve data from views locally or on nearby peers, instead of from the remote sources. Thus network traffic can be reduced sharply and consequently query-processing efficiency can be boosted significantly