A Hybrid Query Scheme to Speed Up Queries in Unstructured Peer-to-Peer Networks

Unstructured peer-to-peer networks have gained a lot of popularitydue to their resilience to network dynamics. The core operation insuch networks is to efficiently locate resources. However, existingquery schemes, for example, flooding, random walks, and interest-basedshortcut suffer various problems in reducing communicationoverhead and in shortening response time. In this paper, we study thepossible problems in the existing approaches and propose a newhybrid query scheme, which mixes inter-cluster queries andintracluster queries. Specifically, the proposed scheme works byefficiently locating the clusters, sharing similar interests withintercluster queries, and then exhaustively searching the nodes inthe found clusters with intracluster queries. To facilitate thescheme, we propose a clustering algorithm to cluster nodes thatshare similar interests, and a labeling algorithm to explicitlycapture the clusters in the underlying overlays. As demonstrated byextensive simulations, our new query scheme can improve the systemperformance significantly by achieving a better tradeoff amongcommunication overhead, response time, and ability to locate moreresources