并行空间连接查询处理  

Parallel Spatial Join Query Processing

作　　者：刘宇[1] 孙莉[1] 田永青[1] 

机构地区：[1]上海交通大学自动化系,上海200030

出　　处：《上海交通大学学报》2002年第4期512-515,共4页Journal of Shanghai Jiaotong University

摘　　要：基于顺序空间连接查询的效率不能令人满意 ,利用并行机制提高空间连接查询效率 .空间连接查询的并行处理方法最重要的特征是任务分配根据多路平面扫描顺序 ,避免了连接处理过程中处理器之间的通信花费 .提出基于空间连接花费模型的任务分配方法和基于花费估计的动态任务分配策略 ,并给出了花费模型 .One of the most important and time-consuming types of query processing in spatial databases is spatial join. The response time of sequential spatial join is far from meeting the requirements of an interactive user. The most important character of the parallel processing method of spatial join query is that the distribution of task is based on plane-sweeping algorithm, which avoids the cost of communication between processor in the join processing. The paper put forward the task-distribution method based on spatial join, and the dynamic task-distributing strategy based on cost-estimation, and presented a cost model. The effect of the model is evident in practice.

关 键 词：空间数据库 R树 多路空间连接 并行空间处理 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]