基于模式覆盖的仿真数据匹配算法  

Simulated Data Matching Based on Schema Covering

作　　者：屈树谦[1,2] 王海波[1] 吕品[1] 郑昌文[1] 

机构地区：[1]中国科学院软件研究所,北京100190 [2]中国科学院研究生院,北京100049

出　　处：《计算机仿真》2012年第12期139-143,共5页Computer Simulation

基　　金：中国科学院知识创新工程方向性项目(yyyj-1125)

摘　　要：研究仿真数据优化匹配传输效率问题,随着分布交互式仿真规模的扩大,如何高效地对仿真数据进行分发和过滤成为一个重要的问题。针对分布交互式仿真领域数据类型固定和属性个数跨度较大的特点,在分析和研究已有事件匹配算法的基础上,提出了模式覆盖集的概念,并设计了一种采用模式覆盖技术的仿真数据匹配算法。改进算法综合考虑了谓词之间的相互关系和订阅条件所属模式之间的覆盖关系,从减少无关的谓词匹配和高效组织订阅条件两方面加速了仿真数据的匹配过程。实验结果表明与谓词计数算法相比,改进算法能较大地提高匹配性能。With the expansion of distributed interactive simulation, the efficient dissemination and filtering of simulated information become important problems. Considering the fixed type of simulated data and the broad span of attribute number in distributed interactive simulation field, and based on the study and analysis of current event matching algorithm, a new structure of schema covering set （SCS） was proposed. A Simulated Data Matching algorithm based on the schema covering was then designed in this paper. By considering the relationship among predicates and the covering relation among subscriptions＇ schemas synthetically, the algorithm accelerates the matching performance by reducing redundant predicates matching and organizing subscriptions efficiently. Finally, experiments show that compared with counting method, the algorithm has considerable improvement in matching performance.

关 键 词：模式覆盖 发布 订阅 事件匹配 仿真数据 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]