文献类型：期刊文献

中文题名：基于改进粒子群优化SVM的多分类入侵检测研究

英文题名：Research on intrusion detection based on improved particle swarm optimization SVM

作者：杨智慧[1];王华忠[1];颜秉勇[1];陈冬青[2]

机构：[1]华东理工大学化工过程先进控制和优化技术教育部重点实验室,上海200237;[2]中国信息安全测评中心,北京100085

年份：2016

卷号：41

期号：3

起止页码：779

中文期刊名：广西大学学报（自然科学版）

外文期刊名：Journal of Guangxi University（Natural Science Edition）

收录：CSTPCD;;北大核心:【北大核心2014】；

基金：国家自然科学青年基金资助项目(51407078)

语种：中文

中文关键词：入侵检测;粒子群;支持向量机;多分类

外文关键词：intrusion detection; particle swarm optimization; support vector machine; multi classification

摘要：针对工控网络数据的高维特性以及攻击方式多样性而导致传统入侵检测算法检测准确率低等问题,采用改进粒子群(PSO)算法优化支持向量机的参数,提出改进的PSO-SVM多分类入侵检测方法。该方法将SVM参数作为改进PSO的粒子,将SVM分类准确率作为PSO的目标函数进行全局搜索以确定SVM的最优参数,建立基于改进PSO-SVM的"一对一"多分类工控入侵检测模型。最后采用密西西比州立大学关键基础设施保护中心提出的工控标准数据集进行仿真实验,结果表明,该算法针对不同的攻击方式的平均检测准确率均能达到90%以上,能够准确识别攻击类型,可为工控系统入侵检测提供有效方法。
For the problem of traditional intrusion detection algorithms with low detection accurate rate which is caused by the high dimensional characteristics of the industrial control network data and diversity of attack patterns, a improved particle swarm optimization （PSO） algorithm which used to optimize the parameters of support vector machine （ SVM）, and an improved PSO-SVM multi classification intrusion detection method is proposed. The support vector machine parameters are op- timized by particle swarm of particles while the SVM classification accuracy is used as particle swarm target function for global search to determine the optimal parameters of SVM. Base on the improved PSO-SVM ＂one to one＂ classification an industrial intrusion detection mode is established. Finally, the simulation experiment is carried on with the latest proposed industrial standard data sets by the Mississippi State University Center for critical infrastructure protection. The results show that the average detection accuracy rate of the proposed algorithm can reach more than 90% for different ways of attacking, and can identify the type of attack accurately. The improved PSO-SVM provides an effective method for the intrusion detection of industrial control system.