公路环境下支持高效更新的时空数据索引结构

SPATIOTEMPORAL INDEX STRUCTURE SUPPORTING EFFICIENT UPDATE IN HIGHWAY ENVIRONMENT

随着定位技术在各个领域的应用普及,降低时空数据索引结构中的更新代价已经成为当前很多实际应用的热点研究问题,如定位服务、交通控制等。提出一种针对公路网络中移动对象的索引结构——NLUR-Tree(Network-based Lazy Update R-tree)。NLUR-Tree不但借助于缓冲插入和缓冲删除来减少更新代价,而且还利用公路网络中对象移动的运动模式来提高更新效率。其次,NLUR-Tree还利用缓冲池来管理内存磁盘空间,在一定程度上解决了移动对象在公路网络分布的不均衡性带来的问题。最后,实验证明了NLUR-Tree要明显优于现有的同类型索引结构。

With the popularised application of positioning technology in various fields, to reduce update cost of spatiotemporal data index structure has become hot research subject in many practical applications, such as positioning services and traffic controls, etc. This paper presents an index structure targeted at the moving objects in highway network NLUR-tree （ network-based lazy update R-tree ）. The NLUR-tree, with the help of buffer insertion and buffer deletion, reduces the update cost and uses movement patterns of the objects moving in highway networks to improve the update efficiency. Besides, the NLUR-tree also makes use of buffer pool to manage memory disk space, and to a certain extent it Solves the problem incurred by the imbalance of moving objects distribution in highway networks. Finally, the experiment reported in the paper proves that the NLUR-tree outperforms the existing index structures with same type.