The Research Of GR~*-tree Spatial Index Method

With the development of geographic information system, computer-aided designand manufacturing(CAD/CAM), remote sensing images, digital earth, mobilelocation-based services, and so on, the researches of spatial database and the indexingtechnologies of spatial database had more and more people’s attention. In particular,GIS is a computer system related to spatial data management and spatial informationanalysis, whose core is data which includes the attribute data and spatial data, and thestorage capacity of spatial data is huge, commonly known as massive data, so itmeans higher requirements for representation, storage, management, and retrievalrequest of spatial data. The indexing technique of spatial database is proposed forsolving this problem, and it is also the core of spatial database applications and thekey technology been used to improve the efficiency of spatial database storage andspatial retrieval performance. So, it directly influenced the systems’ success of spatialdatabases. However, traditional indexing techniques (such as B-trees, B+-tree, etc.)are not effective to spatial data, and dedicated indexing techniques (such as R-tree,quadtree, etc.) proposed only for a single data type, which does not apply to all spatialdata. But as people’s continuous demand is improving, there is an urgent need topropose a better, more convenient, and scalable spatial index. Moreover, the currentspatial data types are various, that is why a single indexing technology cannot meetthe rapid lookup of spatial data. And with the huge increase in the amount of data andthe constantly update of index data, their performance has a sharp decline. Therefore,the indexing researches in regard to the mass objects of different spatial data areimminent, which has been a hot research field of spatial database, and is also thefocus study of this paper.This paper introduced several common mainstream indexing technologies ofspatial database. First, it introduced the basic theory of spatial data and spatialdatabases, the research ideas of spatial index and its influencing factors and purposes.Secondly, this paper stated the detail information of the basic definitions and mainalgorithms about these indexing methods of spatial database, and summarized their respective advantages and disadvantages. Finally, this paper put forword a new spatialindex tree—GR*-tree, and achieved its main algorithm ideas through the code.By comparing with other indexing techniques in the experiment, we found theperformance of this new indexing technology more outstanding.