Premium
Building decentralized resource space on a structured P2P network
Author(s) -
Chen Xue,
Zhang Junfeng
Publication year - 2011
Publication title -
concurrency and computation: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.309
H-Index - 67
eISSN - 1532-0634
pISSN - 1532-0626
DOI - 10.1002/cpe.1621
Subject(s) - computer science , routing (electronic design automation) , semantics (computer science) , resource (disambiguation) , distributed computing , locality , dimension (graph theory) , space (punctuation) , database , theoretical computer science , computer network , mathematics , philosophy , linguistics , programming language , pure mathematics , operating system
Resource Space Model (RSM) is a normalized classification model that uses m ‐dimensional semantic spaces to share and manage various resources in a centralized manner. The Resource Space is a semantic coordinate system that can support SQL‐like multi‐dimensional range queries, but lack automatic construction mechanism and the efficient search algorithm for query routing in the dynamic and decentralized environment. This paper first partitions a centralized Resource Space into a number of decentralized sub‐spaces based on the classification semantics and dimension locality of RSM, and then deploys these sub‐spaces onto a ring‐structured P2P topology HRing that is able to totally preserve the classification semantics and dimension locality of RSM. After that, we build the routing tables, define the dynamic resource operations and propose an efficient routing algorithm for multi‐dimensional range queries on the decentralized P2P Resource Space. Theoretical analysis and simulations show that the construction and maintenance cost, as well as the search performance can reach logarithmic scale. Thus, the decentralized P2P Resource Space can be applied on dynamic and large‐scale resource‐sharing systems. Copyright © 2010 John Wiley & Sons, Ltd.