一种时空数据流中范围聚集查询共享策略

Sharing for Multiple Range Aggregation Queries over Spatio-temporal Streams

根据查询谓词的相似性来实现计算共享是数据流查询优化的主要策略之一,主要针对时空数据流中基于滑动窗口的范围聚集问题进行查询优化,采用时间片段和空间片段分别描述滑动窗口之间和空间范围谓词之间的重叠。提出基于时空分片技术将时空数据流元组分成互不重叠的时空片段。范围聚集查询首先基于时空片段生成中间结果,然后根据查询条件与时空片段的映射关系生成查询结果。实验表明,提出的时空分片策略能大大提高时空数据流中范围聚集查询的性能。

By exploiting query predicate similarities in the data streams,queries optimization is a mainly strategy to share computation.In this paper,based on the sliding windows,an approach is presented for the multiple range aggregation queries over spatio-temporal streams.Time fragments were employed to describe the overlaps between sliding windows and spatial fragments for the overlaps between spatial range predications.Then a technique was proposed to partition the streams into different spatio-temporal fragments without any overlaps.The final query results could be captured from the temporary query results over the spatio-temporal fragments.Experimental result shows that our approach can substantially increase performance of the multiple range aggregation queries over spatio-temporal streams.