摘要
针对目前入侵检测系统不能有效检测未知攻击行为、学习能力较差的问题,根据RBF神经网络的自学习、收敛速度快的特性,将RBF神经网络与入侵检测技术相结合,提出了一种RBF入侵检测模型,并对模型各个组成部分进行了分析,最后采用DAR-PA入侵检测数据库中的部分数据,在Matlab下进行了仿真实验。实验结果表明,此模型具有较高的检测率和较低的误报率,可有效地检测出已知和未知攻击行为,有一定的应用价值。
In order to solve the problems that intrusion detection system at present can not effectively discover the unknown attack hehaviours and is weak in learning, according to the characters of RBF neural network in self-learning and quick convergence rates,an RBF intrusion detection model is proposed in combination of the RBF neural network with intrusion detection technology, and each component of the model is also analyzed. At last, the simulation experiment under Matlab is carried out with the use of partial data in DARPA intrusion detection database. The result of the experiment indicates that the model makes high detection rate and low error reporting rate and is able to detect known and unknown intrusion behaviours effectively,so it has certain applied values.
出处
《计算机应用与软件》
CSCD
2009年第6期278-281,共4页
Computer Applications and Software
基金
黑龙江科技学院引进人才科研启动基金项目(04-23)
黑龙江省教育厅科学技术研究项目(10553089)