Research On P2P Network Based Vector Gegraphic Data Organization And Indexing Technogoy

With the popularity of geographical information systems, more and more users want to accessto spatial on-line services and data. Spatial information on-line services face many challenges.Networks transmission of massive spatial data slows the transmission speed. Additionally, requestsof a large number of concurrent users leads poor performance of services. Current researchsolutions have been limited on server side, which is difficult to improve system performance. Theemergence of P2P networks provides a way to solve these problems above. P2P networks can fullyutilize the computing and service capabilities of clients. Therefore, P2P networks can improve theefficiency of distributed vector geographical data organization and the performance of spatialindexing significantly.In this thesis, the vector data organization method and spatial data indexing technology basedon P2P networks are studied. On the vector data organization, a distributed topology basedrepresentation model of P2P vector geographical data is proposed. The model, which is combinedwith vector data division and union ideas by extending WKT (Well-Known Text) format, improvesthe efficiency data transmission. On spatial indexing technology, a new Hybrid P2P SpatialIndexing Network (HPSIN) is proposed, which combines distributed Quad-Tree with DistributedHash Table (DHT) based Chord networks to maintain both load balance and query efficiency of thesystem. The analysis of queuing networks theory shows that HPSIN is scalable and highperformance. In addition, a P2P based prototype system of vector geographical data on-line serviceis implemented, which is based on JXTA platform and Java technology. The testing result showsthat the system can support high efficient vector geographical services fully. The study conclusionsof this thesis could not only provide new theoretic method and idea to study the key technologies ofvector geographical data on-line services, but also have higher theoretic value and wider appliedforeground.