摘要
应用进化编程自动产生若干条模糊规则以检测各种攻击。在计算机网络中,难于明确划分各种进攻的界限,因此在入侵检测系统中高误警率一直是一个主要的问题,然而利用模糊逻辑,能够有效降低误警率。同时规则的自动产生,也提高了系统的灵活性,降低了对本地网络的依赖性。论文最后给出了测试结果。
This paper proposes one approach for intrusion detection based on fuzzy classifiers and evolutionary programming. In computer network, it＇s hard to define the boundaries of various kinds of attacks, how to reduce the false alarm rate is a mean problem, by fuzzy logic, a set of fuzzy rules can be used to define various kinds of attacks to reduce the false alarm rate. At the same time, this approach builds rules atomically, which makes the system more flexible and independent on local network. Finally the paper exhibits some result of fuzzy classifiers in intrusion detection.
出处
《计算机工程》
EI
CAS
CSCD
北大核心
2005年第23期164-166,共3页
Computer Engineering
基金
综合业务网国家重点实验室开放基金项目(ISN6-7)
国防科技预研基金资助项目
关键词
网络安全
入侵检测
模糊规则
Network security
Intrusion detection（ID）
Fuzzy classify
