基于扩展贝叶斯分类算法的网络安全检测  

Expanding Bayes algorithm in network security model

作　　者：孙亚男[1] 宁士勇[1] 刘文俊[2] 

机构地区：[1]清华大学 [2]中南财经政法大学信息学院,湖北武汉430000

出　　处：《计算机工程与设计》2004年第6期954-957,共4页Computer Engineering and Design

摘　　要：很多网关系统和入侵检测系统被设计来保护自身网络系统的安全,其中一种安全隐患是现有网关系统的不完备性和入侵检测系统的虚警和漏警。总结了前人成果,将贝叶斯算法进行了改进并与之融合,对原有的入侵检测算法提出了改进,得到了一个新的模型。该模型提高了入侵检测系统的完备性和准确性,更有效地保障了网络系统的安全。Although various gateway and intrusion detection system have been designed to protect system security, they work with a hidden trouble on the imperfectness of gateway and the error diagnosis of intrusion detection system. The traditional Bayes algorithm in the intrusion detection system is improved, and the absolute coordinate is used to replace the relative coordinate, so it is easy to see the detected objeet's manner with in the scope of rules or out of the normal behavior. Besides, this model weakens the requent for independence of all the object's behavior and so it improves performance of intrusion detection system and protec-tects the security of information system more effectively.

关 键 词：贝叶斯算法 网络安全 入侵阻止 入侵检测 规则检测 网关系统 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]