Author Login Editor-in-Chief Peer Review Editor Work Office Work

Computer Engineering ›› 2012, Vol. 38 ›› Issue (10): 51-53. doi: 10.3969/j.issn.1000-3428.2012.10.014

• Networks and Communications • Previous Articles     Next Articles

Simple Continues Near Neighbor Chain Query with Predestination Data Chain Size

ZHANG Li-ping 1a, LI Lin 2, LI Song 1a, HAO Xiao-hong 1b   

  1. (1a. School of Computer Science and Technology; 1b. Computing Center, Harbin University of Science and Technology, Harbin 150080, China; 2. TDLTE System Verification Department, Hangzhou R & D Center, Nokia Siemens Networks Co., Ltd., Hangzhou 310000, China)
  • Received:2012-02-27 Online:2012-05-20 Published:2012-05-20

预定数据链规模的单纯型连续近邻链查询

张丽平 1a,李 林 2,李 松 1a,郝晓红 1b   

  1. (1. 哈尔滨理工大学 a. 计算机科学与技术学院;b. 计算中心,哈尔滨 150080;2. 诺基亚西门子通信技术有限公司杭州研发中心TDLTE系统测试部,杭州 310000)
  • 作者简介:张丽平(1976-),女,讲师、硕士,主研方向:空间数据库技术;李 林,工程师、硕士;李 松,副教授、博士;郝晓红,高级实验师
  • 基金资助:
    国家自然科学基金资助项目(60673136, 60903083);哈尔滨理工大学青年科学研究基金资助项目(2011);黑龙江省教育厅科 学技术研究基金资助项目(11551084);黑龙江省自然科学基金资助 项目(F200702, F201134)

Abstract: This paper researches the Simple Continues Near Neighbor Chain(SCNNC) query with predestination data chain size, based on Hilbert curve, the SCNNC_H_SS algorithm is proposed. The redundant data information can be duly deleted and the number of the effective data point is decreased with operation of the algorithm. The redundant computation is avoided. To maintenance and update the SCNNC, the SCNNC_H_CS algorithm is given. Theatrical analysis and experimental results show that when the scale of data set and the chain are great, the algorithm is superior to the methods based on the tree index structure.

Key words: spatial database, spatial data mining, nearest neighbor query, continues near neighbor chain, R tree, Hilbert curve

摘要: 研究预定数据链规模的单纯型连续近邻链(SCNNC)查询问题,基于Hilbert曲线,提出SCNNC_H_SS算法,将已处理过的数据点从数据集中进行剔除,可减少大量冗余计算。为对SCNNC进行动态维护和更新,提出SCNNC_H_CS算法。理论分析和实验结果表明,在数据集和待查近邻链的规模较大时,相比基于传统树索引结构的方法,该算法具有更高的查询效率。

关键词: 空间数据库, 空间数据挖掘, 最近邻查询, 连续近邻链, R树, Hilbert曲线

CLC Number: