Research On Cache Mechanism For P2P Based Spatial Vector Data Indexing Network

Geographic Information Systems takes the form of data to reveal the objects of the real world(roads, land use, elevation, etc.). How to obtain the data quickly and easily from the vast amounts ofdata required by the user has become the focus problem. Spatial indexing technology provides amethod to increase the indexing speed for the spatial data while the indexing performance woulddecrease with the increase in the amount of spatial vector data. The spatial vector data indexingnetwork based on P2P can effectively make use of the edge of network resources and achieve datasharing rapidly.In this thesis, a caching mechanism is proposed based on analyzing and summarizing theexisting hybrid P2P spatial indexing network, which can optimize the resources of the nodeselection strategy and give preference to local resources. In allusion to the multi-layer networkstructure and the requirement of user queries, the caching mechanism can store the multiple layersof spatial vector data. In addition, a novel cache replacement strategy for space vector data ofmulti-layer is proposed. It’s abstracted as a mathematical model of the0/1Knapsack problem andsolved by genetic algorithms. The simulation experiments indicate that this strategy can increase thecaching hit ratio and improve the indexing efficiency effectively. Finally, a prototype system forspatial vector data indexing network based on P2P is implemented. The experimental resultsindicate that the indexing network with the caching mechanism can make full use of local resourcesand reduce the response time effectively. The research results of this thesis improve the quality ofonline service system for spatial vector data, and also provide new ideas for mass data indexingnetwork.