摘要
为了检测无线局域网MAC层的6种DOS攻击方式,提出一种基于Hybrid特征选择和支持向量机的入侵检测算法。该算法先用混合器模式的Hybrid特征选择算法提取8个识别攻击的流量统计特征,然后利用支持向量机对待检测对象进行识别分类。通过建立仿真环境对检测模型的检测效果进行统计验证,表明检测模型在具有较高检测准确率和较低的虚警率,能够有效地检测MAC层DOS攻击,具有实用价值。
A new intrusion detection algorithm based on Hybrid feature selection is proposed to detect denial of service (DOS) attack in media access control(MAC) layer of WLAN. Firstly, this method uses Hybrid featureselect algorithm to collect 8 traffic statistical features to distinguish attacks, and then it uses support vector machine support vector machine(SVM) to classify the test data. Accordingly, the simula- tion environment is built to verify the test efficiency. The experimental result shows that the scheme makes a high detection rate and low false alarm. It is effective to test the DOS attacks in MAC layer.
出处
《现代防御技术》
北大核心
2013年第3期75-80,105,共7页
Modern Defence Technology
关键词
无线局域网
介质访问层
拒绝服务攻击
特征选择
支持向量机
入侵探测
WLAN
media access control (MAC) layer
denial of service (DOS) attack
feature selection
support vector machine (SVM)
intrusion detection