Research On A Dynamic Management For Spatial Vector Data Based On P2P Network

With the increasing of the amount of space vector data and the growing of computationalcomplexity, the centralized mode of GIS encounters many problems. P2P network as a newdistributed computing model can use the source of the edge node effectively to accomplish the datastorage and computing operations. The hybrid P2P spatial indexing network which combined P2Pnetwork and GIS technology has the potential to solve the mass data storage problem and canalleviate the bottleneck effect of data index caused by the concentrator in centralized GIS.Currently, the nodes in the network are not stabile enough and likely to cause data missing andindex disordering. In order to solve this problem the dynamic management for spatial vector databased on P2P network is promoted in this paper. This management can accomplish the vector data’sredundancy backup, recovery data or update index with the backup data. It is effectively ensure thedata’s integrity and index’s correctly. What’s more, the index efficiency can be further improved byoptimizing the routing mechanism. PeerSim simulation was used to verify the feasibility andperformance advantages of the management, and at the last of this thesis we use JXTA platformrealizing the prototype system of the dynamic data management mechanism.Theoretical analysisand experimental results show that the dynamic management can accomplish the spatial vectordata’s backup and recovery, ensure the integrity of data. It is also can improve query success rateand shorten the query latency.