摘要
空间连接运算是空间数据查询中最重要、最耗时的基本操作之一,其中基于R树的空间连接(RJ)被认为是一种高效的处理机制,但在空间连接的精化阶段处理复杂的空间数据时需要很大的系统开销。基于MBR及直接查询谓词,提出了一种加权处理方法,并扩展了R树结构及MRJ算法。从而优化了多路R树连接的筛选处理,能得到更加有效的候选集;同时,减少了磁盘访问次数,可节省CPU及I/O的时间开销。还通过应用实例验证了其在空间数据库查询优化方面的优势。
Spatial join is the most important and time-consumed in spatial query of SDBMS,R-tree based spatial join is always considered as a kind of high efficient processing mechanism.In the refinement step of spatial join,it needs a lot of system overhead to process complex spatial data.This paper based on the MBR and the direct query predicate proposes a method to generate a weighted complete query graph and extends the structure of R-tree and the MRJ algorithm which mainly contributes to optimize the filter step of the multi-way R-tree join to get a set of more effective candidates and then reduce the frequency of disk access and the overhead of CPU as well as I/O.Through the experiment of specific application by using this method,we show the significant superiority in optimizing the spatial query of SDBMS.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第31期174-178,共5页
Computer Engineering and Applications
基金
信息产业部资助项目(2003xk320014)
关键词
空间连接
多路R树连接
派生谓词
查询图
加权处理
spatial join
multi-way R-tree join
derived predicate
query graph
weighted processing