Research On Spatial Data Query For Peer-to-Peer Environments

As the development of spatial data acquiring, network, spatial database technology, etc, spatial data resources have been fully enriched. How to effectively manage and organize these geographically wide-distributed mass spatial data resources, then finally setup a large scale spatial data shared platform for GIS has become the tough problem need to be solved. However, architecture with the combination of centralized control and partial autonomy brings single location invalidation, performance bottleneck, and inextensibility, which baffles the large scale spatial data resources sharing.Therewith, peer-to-peer computing(P2P for short) technology which is deemed to the key technology for the reconstructing distributed architecture in the future, has been soaringly developing in recent years, and achieved great success in multiple application fields including file sharing, instant communication, streaming media transmission and distributed computing, etc. That P2P system put up advantages of scale, availability, scalability, etc, has informed us that organizing and managing spatial data resources by P2P technology for setting up large scale data shared platforms is a good choice and the way to be developed.The decentralization, high-dynamics, large-scale, of P2P system and complexity of spatial data make a lot of problems for the maturity of spatial data management utilization under the P2P umbrella. On account of the importance of query for data management, this paper focuses on the problem of spatial query under the P2P umbrella, which aims at query support for P2P spatial data management environment.This paper contributes to the following four facets:(1) Applying P2P service system to the P2P data management, construct a querying-oriented P2P spatial data management application framework, and put forward a spatial data resource architecture in this framework.(2) Traditional distributed spatial data indexing method is hard to apply in the P2P environment. This paper argues a distributed spatial indexing method combined with hierarchical-based ring geometry overlay and centralized spatial data indexing technology that is described from several aspects like the construction of index, and the mapping among nodes. (3) Described in detail that how to use spatial point Query, range Query, area Query and KNN Query to locate on the group lay, and put forward the methods of second location on peer lay, which are used to reduce query cost effectively.(4) Aiming at the possible node invalidation in hierarchical geometry overlay, this paper detailed analyzed the triggered problems due to invalidation, and put forward the routing recovery methods in hierarchical geometry overlay based on space taken, which are used to maintain reply for active peers in the system.