摘要
针对分布式并行环境下海量空间数据管理与并行化处理的效率问题,以提高分布式并行空间数据的查询效率为目的,根据现有的空间索引结构与并行化技术,提出一种新的分布式并行空间索引结构——DPR树。DPR树是空间索引技术与并行化技术优化结合的成果。DPR树在数据的总体划分与部分查询中所采用的均是基于高效处理技术。它在原有的并行Master-client R树的基础上进行改进,采用了HCSDP数据划分技术,并将其应用到分布式环境下,且每个节点机中各子树采用了改进的R树——R*Q树。通过性能分析表明,该索引结构具有高效的查询性能。
On account of the problem of mass spatial data manage ment and parallelizable efficiency which in distributed parallel environment and in order to improve the querying efficiency,a new struct of distributing parallel spatial index struct-DPR tree is offered which based on existed spatial index struct and parallelizable technologies.DPR tree is spatial index struct and parallelization technology optimized composed.DPR tree adopts high efficiency processing technique both on the total divide and partial query.It is based on parallel Master-client R tree and improves,at the same time,introducing HCSDP division technology,and applying it to distributed environment,moreover,every subtree uses the improving R tree—R＊Q tree.By performance analysis indicate that the index struct has efficient query performance.
出处
《计算机技术与发展》
2010年第6期39-42,共4页
Computer Technology and Development
基金
黑龙江省自然科学基金资助项目(F2007-01)
