基于改进蚁群聚类的入侵检测算法  

Intrusion detection algorithm based on improved ant colony clustering

作　　者：杜少波[1] 岳珍梅[2] 

机构地区：[1]河南科技大学电子信息工程学院,河南洛阳471003 [2]许昌职业技术学院信息工程系,河南许昌461000

出　　处：《兰州理工大学学报》2013年第5期85-88,共4页Journal of Lanzhou University of Technology

摘　　要：针对蚁群聚类算法存在容易出现停滞现象和过早地收敛于局部最优解的问题,提出一种改进的蚁群聚类入侵检测算法.通过改进蚂蚁搜索解的方法,来改善蚁群算法易于过早地收敛于非最优解的缺陷.使用KDD99作为入侵检测数据集进行仿真实验,结果表明,改进的蚁群聚类算法能有效提高入侵检测的检测率和降低误检率.Aimed at the problems in ant colony clustering such as ready at stagnation phenomenon and early converging to local optima, an intrusion detection algorithm was presented based on improved ant colony clustering. By means of improving the method for ant searching solution, the early converging to local optimum solution was avoided. The KDD 99 intrusion detection data were used to conduct simulation experiment and the simulation results show that the improved ant colony clustering could effectively improve the detection rate of intrusion detection and reduce the false detection rate.

关 键 词：聚类分析 蚁群聚类算法 信息素 检测率 误检率 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]