基于变精度粗糙集的入侵检测技术研究

Intrusion Detection Research Based on Variable Precision Rough Set

随着网络攻击技术层出不穷,网络攻击方式呈现出多样性和隐蔽性的特征,网络入侵检测的漏报和误报频发。为了提高入侵检测的检测率、降低漏报率和误报率,利用变精度粗糙集挖掘数据之间潜在规律的特性,对测试数据集进行特征提取和属性约简,除去冗余信息降低属性维度,然后通过遗传算法进行分类和检测。实验测试表明该方法不仅可以提高检测率,并能较实时地检测其他类型的攻击。

As the network attack technology after another, showing a diversity of network attacks and covert features, network intrusion detection of false negatives and false alarm frequently. In order to improve the detection rate of intrusion detection, reducing the false negative rate and false alarm rate, the use of variable precision rough set of potential characteristics between data mining law, the test data set for feature extraction and attribute reduction, removing redundant information to reduce the dimension attributes and then classify and detect genetic algorithm, test results obtained. Experimental results show that this method can improve the accuracy of a denial of service attack detection type, and can be detected more complete other types of attacks.