基于主成分分析的实时网络入侵检测仿真  

Emulation for real-time network intrusion detection based on Principle Component Analysis method

作　　者：谢蕾[1] 谢华[2] 

机构地区：[1]信阳师范学院网络信息与计算中心,河南信阳464000 [2]信阳职业技术学院数学与计算机系,河南信阳464000

出　　处：《微计算机信息》2010年第16期166-168,共3页Control & Automation

摘　　要：对网络连接数据进行预处理和特征提取,将其转变成适合于入侵检测的数据集。采用主成分分析(PCA)方法对海量网络连接数据集进行降维,在低维子空间进行入侵检测。建立了仿真模型,仿真结果表明,该方法计算量小,处理速度快,实现了对海量网络入侵数据的实时检测。Through pretreatment and relevant feature selection, the network connecting data can be transferred to some data sets which are suitable for intrusion detection. A intrusion detection emulation model is developed based on the principle component analy- sis（PCA） method to monitor computer network behaviors, Testing results show that the proposed modes is promising in terms of detection accuracy, computation expense and implementation for real-time intrusion detection.

关 键 词：入侵检测 网络安全 主成分分析 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]