Cloud Computing Environment Index Of Gml Spatial Data Storage Mechanism Research

GML is an XML-based encoding of spatial data, widely accepted by GIS fields andenterprises.GML is an important part of cyberspace information. With the development ofinformation technology, the amount of information space is increasing sharply, how GML spatialdata for efficient storage and management has become the focus of research. According to theown characteristics of GML spatial data (distribution of geographic), the use of centralizedstorage is not conducive to the storage of spatial data management and access. The cloud storageis a commercial implementation of distributed storage; It is some of the idle resources ofinexpensive computers together, sharing the tasks of massive data storage and management. Sothe use of distributed storage and management of GML spatial data is a good program.How storage effectively for the mass GML spatial data, mainly for the following aspects ofresearch and innovation analysis:First, for a reasonable division of spatial data, we should consider the multidimensionalnature of spatial data, also consider the spatial relationships and topological relations; Based onthe analysis of spatial data over several division, hierarchical decomposition algorithm of theHilbert curve and spatial data divided of GML vector layers has been studied. Presented tomethod of the division of GML spatial feature layers combined with rapid coding of Hilbert curvehierarchical division into processing spatial elements objects. In specific applications, refer to thestorage node information divide the elements objects into sub-blocks, and assign the appropriateidle resources for these distributed blocks.Secondly, in the cloud computing environment designed the distributed database. Combinedwith the characteristics of GML spatial data and the master-slave architecture of Hadoop platform,the storage-related table structures and data types were designed on each node. Indexingmechanism of GML spatial data distributed storage and the structure of indexing table have beenresearch and analyzed.The main studied is the characteristics of the QR tree and build an index ofGML spatial data.Finally, in the master-slave architecture of Hadoop open-source platform design theprototype system of distributed storage and partial implementation. And made a comparativeanalysis about efficiency of stand-alone storage and distributed storage; obtained to vast amountsof GML spatial data, whether in storage or query access, distributed fleet processing efficiency ishigher than the stand-alone processing efficiency.Innovation mainly through the Hilbert curve of GML spatial data using classificationmethod of the elements queue；Combined with hadoop platform to create parallel spatial indexingmechanism of GML data and design the index table structure of the child nodes；To design the distributed storage processes of GML spatial data；cloud storage system functional design andrealization.