基于模糊C均值算法的入侵检测方法  

An Approach for Intrusion Detection Based on Fuzzy C-means Algorithm

作　　者：林荣亮[1] 张文波[1] 

机构地区：[1]中国人民解放军92854部队

出　　处：《计算机与数字工程》2012年第5期82-83,148,共3页Computer & Digital Engineering

摘　　要：聚类分析是一种有效的异常入侵检测方法,可用以在网络数据集中区分正常流量和异常流量。采用模糊C均值聚类算法对网络流量样本集进行划分,从中区分正常流量和异常流量,并针对入侵检测问题的特性提出了聚类中心确定方法。最后,利用KDD 99数据集进行实验,证明该算法能够有效地发现异常流量。Clustering analysis is an effective method of anomaly intrusion detection, which can find normal flow and abnormal flow in the network data set. Fuzzy C-means clustering algorithm is applied to classify the network traffic data into normal flow and abnormal flow. A new clustering center method which is designed for intrusion detecting problem specially is provided in this paper. Finally, KDD 99 data set is used for the illustrative example, and the result proves that this algorithm could discover the abnormal flows effectively.

关 键 词：模糊聚类 入侵检测 距离测度 模糊C均值 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]