基于人工免疫分类器的入侵检测方法

Intrusion Detection based on Artificial Immune Classifier

针对入侵检测系统准确率不高和难以检测未知攻击的缺点,将有限资源人工免疫分类器模型算法AIRS应用于入侵检测系统。首先从KDD CUP 99数据集中选取出部分正常数据和攻击数据,对AIRS算法进行训练。然后根据训练得到的模型,对包含已知攻击和未知攻击的不同异常类比的数据集进行测试。实验结果表明:AIRS算法对已知攻击的检测率大大提高,对未知攻击的识别率也有很大的提高。

Based on the shortcomings of intrusion detection system in which accuracy rate is not high and unknown attacks is difficult to be detected, a resource limited artificial immune classifier model algorithm AIRS is used in the intrusion detection system. A part of the normal data and attack data are selected from the KDD CUP 99 data set for training AIRS algorithm at first. Then the trained model is used to test data set with different abnormal analogy which contains known attacks and unknown attacks. The experimental results show that the AIRS algorithm has greatly improved the detection rate of known attacks and the recognition rate of unknown attacks.