摘要
分布式入侵检测系统需具有分布式检测功能及部件增量更新能力。文中提出了一种基于神经网络集成的分布式入侵检测方法,采用单个Agent检测与多个Agent协同检测的两级集成算法实现分布式入侵检测;在发现新的入侵时,Agent上的神经网络集成采用基于资源分配网的增量学习算法进行更新。实验结果表明,该算法能有效检测各种攻击,并且具有对未知攻击的增量学习能力。
Distributed intrusion detection system requires abilities of distributed detection for intrusions and incremental update for its components. A novel distributed intrusion detection method based on neural network ensemble is proposed. The distributed detection is implemented by a ranked ensemble algorithm. It is firstly detected in single agent with an ensemble of neural networks and then is cooperated with other agents to obtain detected outcome while one agent cannot detect by itself. When discovering a new kind of attack, neural network ensemble is updated by a resource allocating network （RAN） based incremental learning algorithm. Experimental results show that the algorithms are effective in detecting attacks.
出处
《南京航空航天大学学报》
EI
CAS
CSCD
北大核心
2007年第2期231-235,共5页
Journal of Nanjing University of Aeronautics & Astronautics
基金
江苏省自然科学基金(BK2005135)资助项目
关键词
分布式入侵检测
神经网络集成
增量学习
攻击
distributed intrusion detection
neural network ensemble
incremental learning
attack
