It is necessary to build spatial index in order to managing complex spatial data in peer-to-peer (P2P) environments. This
paper analysis and summarizes the related studies, designs a new spatial index model named P2PQR-tree which uses the
distributed Quad-tree and the local R*-tree. P2PQR-tree applies Quad-tree techniques into P2P environments and uses
replication strategy to improve the system's load balance. This paper includes gives the models definitions, the model's
architecture design, the model's performance analysis results. P2PQR-tree has some advantages to the old methods, for
example, its data management is more reasonable, it can support metadata management better, implement rights control
easier, reduce changes of distributed index, and adapt to dynamic character of P2P network better.
|