基于R树的空间查询连接处理优化与实现  

Optimization and implementation of R tree-based processing on spatial join query

作　　者：吕闽晖[1] 吕敏蓉[2] 

机构地区：[1]海军工程大学装备经济研究所,湖北武汉430033 [2]湖南女子学院,湖南长沙410004

出　　处：《微型机与应用》2011年第13期66-68,共3页Microcomputer & Its Applications

摘　　要：空间索引作为空间数据库的关键技术,其性能的高低决定着整个空间数据库的效率。通过对现有的多种空间索引结构进行比较分析,基于开源数据库Ingres实现了广度优先R树连接算法(BFRJ),并对其进行了局部优化和全局优化。基于真实数据的实验结果分析,证实了采用适当的全局优化方法的BFRJ优于其他已知的空间连接算法方法。As the key technique of spatial database, the performance of spatial index determines the efficiency of the entire database. This paper works on the open-source database named Ingres. Through a comparative analysis of existing spatial join structures, we propose to implement the Hilbert R-tree. Then, we implement the Bread-First R-tree Join Algorithm （BFRJ）, and provide local and global optimazition. At last, through a series of experiments based on the real world data sets, we comfirms that, with the appropriate choice of global optimazition, the BFRJ is better than other spatial join algorithms.

关 键 词：空间索引 空间连接 HILBERT R树 BFRJ 

分 类 号：G304[文化科学]