基于路网的移动对象动态双层索引结构  

A Two-Tiered Dynamic Index Structure of Moving Objects Based on Constrained Networks

作　　者：乔少杰[1] 韩楠[2] 王超[1] 祝峰[3] 唐常杰[4] 

机构地区：[1]西南交通大学信息科学与技术学,成都610031 [2]西南交通大学生命科学与工程学院,成都610031 [3]闽南师范大学福建省粒计算及其应用重点实验室,福建漳州363000 [4]四川大学计算机学院,成都610065

出　　处：《计算机学报》2014年第9期1947-1958,共12页Chinese Journal of Computers

基　　金：国家自然科学基金(61100045,61165013);高等学校博士学科点专项科研基金(20110184120008);中国博士后科学基金特别资助项目(201104697);教育部人文社会科学研究青年基金(14YJCZH046);中央高校基本科研业务费专项资金(2682013BR023)资助~~

摘　　要：为了支持对大规模不确定性移动对象当前及将来位置的查询，亟需设计更加有效和高效的索引结构．当前索引算法主要考虑索引建立和维护的效率问题或关注基于索引进行查询时的准确性，对索引建立维护以及查询时性能综合考虑的研究较少．针对已有方法的不足，提出基于路网的移动对象动态双层索引结构DISC-tree，对静态路网信息采用R＊-tree索引，对实时更新的移动对象运动轨迹采用结点更新代价较小的R-tree进行索引，设计哈希表和双向链表辅助结构对索引协同管理．成都市真实地图数据集上的实验结果表明：相比于经典的NDTR-tree，DISC-tree在索引建立和维护方面时间代价平均减少39．1％，移动对象轨迹查询时间代价平均减少24．1％；相比于FNR-tree，DISC-tree的范围查询准确率平均提高约31．6％．In order to support efficient query for the current and future position of massive and uncertain moving obiects, it urgently needs to propose effective and efficient index structures. The existing algorithms focus on the efficiency of creating and maintaining the index structure or the accuracy of index-based query processing, and the algorithms which take into consideration the performance of both index maintenance and query performance are rarely reported. Aiming to handle the drawbacks of existing algorithms, a new index called DISC-tree is proposed, which uses R＊ -tree to manage the static road networks and employs low-cost R-tree to index trajectories of moving objects that update in real time. Moreover, an associated structure integrates hash map and doubly linked list is applied to index dynamic trajectories. The experiments are conducted on synthetic datasets generated on the real map of Chengdu, and the results show that the DISC-treecan reduce the time cost of index creation and maintenance by an average gap of 39.1% compared with the classical NDTR-tree, as well as reduce the time cost of trajectory query by an average gap of 24.1%. In addition, compared to the FNR-tree, the accuracy of range query is averagely improved by about 31.6%.

关 键 词：移动对象数据库 路网 索引 范围查询 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]