Parallel indexing technique for spatio-temporal data
Parallel indexing technique for spatio-temporal data
复制标题
时空数据并行索引技术
DOI：
10.1016/j.isprsjprs.2013.01.014
复制
发表时间：
2013-04-01
影响因子：
12.7
通讯作者：
Xiao, Jing
中科院分区：
文献类型：
--
作者：
He, Zhenwen;Kraak, Menno-Jan;Xiao, Jing
在地理信息系统及其应用中有效访问和管理大规模多维时空数据的要求得到了充分认可和研究。最流行的时空访问方法是R-Tree及其变体。但是，很难将它们并行访问多维时空数据，因为R-Trees及其变体是在高维空间中存在严重重叠问题的层次结构中。我们将间隔的二维间隔空间表示扩展到了多维并行空间，并提供了一组公式，以将时空查询转换为并行的间隔设置操作。这种转换将多维对象关系的问题降低到更简单的二维空间交集问题。实验结果表明，本文介绍的新的并行方法比处理多维时空数据和多维间隔数据的R*-Trees具有出色的范围查询性能。当CPU核心的数量大于空间维度的数量时，这种新方法的插入性能也优于R*-Trees。所提出的方法为快速数据检索大规模的四维或更高维数数据提供了潜在的平行索引解决方案。 （c）2013年国际摄影和遥感学会（ISPRS）由Elsevier B.V.保留所有权利。
The requirements for efficient access and management of massive multi-dimensional spatio-temporal data in geographical information system and its applications are well recognized and researched. The most popular spatio-temporal access method is the R-Tree and its variants. However, it is difficult to use them for parallel access to multi-dimensional spatio-temporal data because R-Trees, and variants thereof, are in hierarchical structures which have severe overlapping problems in high dimensional space. We extended a two-dimensional interval space representation of intervals to a multi-dimensional parallel space, and present a set of formulae to transform spatio-temporal queries into parallel interval set operations. This transformation reduces problems of multi-dimensional object relationships to simpler two-dimensional spatial intersection problems. Experimental results show that the new parallel approach presented in this paper has superior range query performance than R*-trees for handling multi-dimensional spatio-temporal data and multi-dimensional interval data. When the number of CPU cores is larger than that of the space dimensions, the insertion performance of this new approach is also superior to R*-trees. The proposed approach provides a potential parallel indexing solution for fast data retrieval of massive four-dimensional or higher dimensional spatio-temporal data. (C) 2013 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS) Published by Elsevier B.V. All rights reserved.