数据挖掘在光纤网络入侵检测中的应用

Application of data mining in optical fiber network intrusion detection

光纤网络入侵检测一直人们关注的焦点,针对光纤网络入侵数据庞大的特点,利用粗糙集属性约简和支持向量机优异的分类性能,提出一种数据挖掘的光纤网络入侵检测模型。首先对光纤网络数据进行归一化处理,消除特征单位不一致带来的干扰,然后采用粗糙集对光纤网络入侵检测样本的特征进行约简,降低光纤网络特征的数量,最后采用支持向量机建立光纤网络入侵检测模型,并通过KDDCUP99数据集进行仿真实验。结果表明,本文模型可以从历史数据中挖掘光纤网络入侵变化特点,提高了光纤网络入侵检测正确率,并且执行效率高,可以达到了光纤网络入侵检测的预期目的。

Optical fiber network intrusion detection has been the focus of attention, for the characteristics of optical fiber network intrusion data, using rough set attribute reduction and excellent classification performance of support vector machine, a new optical fiber network intrusion detection model based on data mining is proposed. First, the data of the optical fiber network data is normalized to eliminate the interference caused by the different feature units, and then we use rough set to reduce the characteristic of the fiber optic network intrusion detection sample, and reduce the amount of optical fiber network data, finally, the support vector machine is used to establish the optical fiber network intrusion detection model, and the simulation experiment is carried out through the KDDCUP99 data set. The results show that the model in this paper can be used to mine the characteristics of optical network intrusion in the historical data, the accuracy of intrusion detection is improved and the execution efficiency is high, and the expected purpose is achieved.