Esearch On Query And Optimization Technology To The Location Based Serrvice Below The Cloud Environment

With the rapid development of information and communication technology，computing mode has experienced from the initial task focused delivered to largeprocessor mode， latterly, development for distributed task processing mode basedon the network， and then to the on-demand treatment of cloud computing modelCloud computing as a revolutionary computing model has become important inmany industries technology trends. The space cloud computing cloud computing applications are formed in geographic information field has gradually become themainstream technology of spatial geographic information industry. In recent years， along with the location based services (LBS)， spatial data is the rapid growth， the spatial data index mechanism brings a great impact， and the index of these traditional methods are often based on memory or a prerequisite to optimization of disk access. Therefore，how to achieve efficient spatial index and query processing in large-scale spatial data has become the new demands and challengesof cloud computing applications. The best choice for spatial data in a scalable，distributed query technique is efficient spatial data query and analysis.This paper designs the reverse mesh indexing service to the position of a cloud environment below and parallel KNN query. The main work:(1) introduced the index structure of several spatial data, such as R-tree index,spatial grid index and the spatial index based on Voronoi diagram, and analyzesthe existing problems of these index structures; the spatial KNN query algorithm,and analyzed the advantages of this kind of query algorithm and the existing problems.(2) the related concepts, the key technology of cloud computing are briefly introduced. Due to the reverse grid index has the advantages of simple structure,fast updates, easier to handle high-dimensional data characteristics, this paper presents a reverse grid index based on MapReduce programming model.(3) because of the complexity of spatial data, and considering the efficiencyindex, this paper adopts the grid index to establish the index, and propose the MapReduce grid based indexing method. Because the grid indexing and query algorit hms are independent of each other, namely an indexing, multiple query processing, this paper presents a parallel KNN algorithm based on reverse grid index.(4) To build the cloud computing platform，the establishment of reverse grid indexand KNN parallel query. The experimental results show that， the proposed indexing time is significantly lower than the R-tree and the Voronoi polygon index time， at the same time， spatial query on the index is better than the other two index.