并行R树空间索引中叶节点大小的确定方法研究  

Research on the method for evaluating the size of leaf node of parallel R-tree spatial indexing

作　　者：赵园春[1] 李成名[2] 赵春宇[3] 

机构地区：[1]山东科技大学地球信息科学与工程学院,山东青岛266510 [2]中国测绘科学研究院,北京100039 [3]武汉大学遥感信息与工程学院,武汉430079

出　　处：《测绘科学》2008年第4期94-97,共4页Science of Surveying and Mapping

基　　金：科技部863项目(2001AA136010)

摘　　要：并行R树空间索引结构中叶节点的大小是影响索引效率的主要因素,其确定方法是并行R树索引结构性能优劣的关键。本文讨论并设计了一种多层并行R树空间索引结构,文中以系统的查询响应时间作为性能评估指标,给出了并行R树叶节点大小的确定方法,并通过实验验证了该方法的有效性和适用性,同时也论证了本文所设计的多层并行R树索引结构是合理的和高效的。The size of leaf node of parallel R-tree spatial indexing is one of the important factors for the efficiency of spatial indexing, and the evaluation method will determine the stand or fall of the performance of parallel R-tree spatial indexing. The paper discusses and designs a multi-tiers parallel R-tree spatial indexing structure under the distributed parallel computing environment, and using the response time of system query as the performance evaluation factor, gives the method for evaluating the size of leaf-node of parallel R-tree spatial indexing, and the availability and the applicability of the evaluation method has been validated sufficiently by experimenting. Simultaneously, the experiment proves that the structure of parallel R-tree spatial indexing is rational and high-performance.

关 键 词：并行空间索引 并行R树空间索引 并行GIS 分布式并行计算环境 

分 类 号：P208[天文地球—地图制图学与地理信息工程] TP391[天文地球—测绘科学与技术]