基于CPS的协议识别优化算法  

Protocol Identification Optimization Algorithm Based on CPS System

作　　者：王健[1] 栾方军[1] 邱海斌[2] 王长涛[1] 付佳新[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168 [2]沈阳广播电视大学,辽宁沈阳110003

出　　处：《科技广场》2012年第8期62-66,共5页Science Mosaic

基　　金：辽宁省自然科学基金项目(编号:20072009);住建部科研项目(编号:2010-K9-49)

摘　　要：目的:自美国将CPS的发展列为重点研究项目,引起了各国广泛的关注。不同于传统的系统,在通信方面,CPS融合了大量的异构网络,其规模远大于现有的控制网络。为提高对协议识别的效率和准确度,并对其进行分类,系统中大量数据传输带来的网络拥堵是我们要解决的主要问题。方法:基于网络流的特征,提出了对协议进行识别的变精度粗糙集的权值计算法和决策构造优化算法。结果:基于变精度粗糙集算法协议识别算法,其准确率可达98.7%。结论:基于变精度粗糙集的决策算法相较于普通决策算法,其识别准确度更高、更容易实现。The United States has ranked the development of CPS as a key research project that has caused various countries＇ wide attention.Different from the traditional system,in communications,CPS integrated a lot of heterogeneous network,whose scale of CPS far outweighs the existing control network.In order to improve the efficiency and accuracy of the recognition and to realize network protocol classification,network congestion that is resulted by transmitting a large number of data in the communication system is the main problem to be solved.Proposing weight calculation based on variable rough set and decision optimization algorithm that both are based on characteristics of network flow in protocol identification.Protocol identification algorithm based on variable precision rough set can make its accuracy rate up to 98.7%.The results show that,with ordinary decision-making algorithm,decision-making based on variable precision rough set has higher reorganization accuracy and it＇s easier to achieve.

关 键 词：CPS系统 协议识别 变精度粗糙集 决策树 

分 类 号：TN915.04[电子电信—通信与信息系统]