面向动态环境的移动对象自适应索引方法  

Self-tuning indexing of moving objects for dynamic environment

作　　者：陈楠[1] 寿黎但[2] 陈刚[2] 陈珂[2] 胡天磊[2] 

机构地区：[1]浙江省烟草公司信息中心,浙江杭州310001 [2]浙江大学计算机科学与技术学院,浙江杭州310027

出　　处：《浙江大学学报（工学版）》2013年第3期442-448,共7页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(60603044;60803003);国家"863"高技术研究发展计划资助项目(2009AA01Z137);浙江省科技计划资助项目(2008c141060)

摘　　要：为适应动态变化的应用环境,提高索引的综合性能,提出一种自适应的移动对象索引———STBx树.给出了STBx树的索引结构、更新算法和查询算法,并且在性能分析的基础上给出了STBx树进行自适应调节的方法.STBx树以自学习和自适应的运行,在不打断服务的情况下对自身的更新性能和查询性能进行调节,从而达到最佳的平稳的综合性能,以适应更新操作和查询操作的比例以及性能需求动态变化的环境.实验表明:STBx树在动态应用环境下能够实现自适应的调节,并提供优秀的综合性能,优于传统的TPR*树和Bx树.A novel index of moving objects based on Bx-tree,named STBx-tree was proposed in order to adapt to dynamic environment and improve overall performance.The index structure,update algorithm and query algorithm of STBx-tree were described.In addition,the self-tuning technique of STBx-tree was designed and represented based on the performance analysis.STBx-tree adapts to dynamic environment in a self-learning and self-tuning way without interrupting the indexing service.STBx-tree adjusts its update and query performance to obtain optimal and smooth overall performance,when the proportion of query and update operations in the application dynamically changes over time.Experimental results show that STBx-tree worked well and efficiently in dynamic environment,and outperformed the existing indexes of TPR＊-tree and Bx-tree.

关 键 词：移动对象 数据库 STBx树 索引 动态环境 自适应 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]