研究网络安全问题,由于网络受到各类攻击,当前入侵检测系统的最大问题是不能快速检测出新出现的异常入侵和较高的误报率.为了提高网络入侵检测正确率,提出一种人工免疫的网络入侵检测算法.网络入侵检测算法模拟了人工免疫系统中的匹配、否定选择、记忆等机制,通过人工免疫的“自我”和“非自我”识别能力对网络入侵行为进行检测.最后在Matlab平台上,对KDD CUP1999网络入侵数据集进行验证性测试,实验结果表明,提出的算法简单,检测准确,能识别出未知入侵信息,提高了检测效率,为设计提供了依据.%Intrusion detection is the research hotspot in network security. Intrusion detection system cannot fast detect new abnormal invasion and the rate of false positives is higher. In order to improve the detection accuracy of network intrusion, a network intrusion detection algorithm is put forward based on artificial immune network. Artificial immune algorithm has adaptability, robustness and good characteristics, "self" and "nonself" are reconditioned by artificial immune algorithm. Finally the proposed method is tested by the KDD CUP1999 network intrusion datasel in Matlab, experimental results show that the proposed method is simple and accurate, and can identify new detection intrusion and improve the detection efficiency.
展开▼
