基于粗糙集理论和BP神经网络入侵检测模型

Model of Intrusion Detection Based on Rough Sets and BP Neural Network

研究网络入侵准确检测问题。针对入侵检测系统存在的比较高的漏报率以及高的误报率,同时也存在入侵检测的数据存在维数大、冗余度高等缺陷。为了保证网络的安全防护技术的实时性和有效性,结合领域粗糙集和BP神经网络算法的优点,提出了一种新的基于领域粗糙集理论和BP神经网络算法的入侵检测算法。首先在粗糙集理论的基础上引入领域概念,减少信息的丢失,利用领域粗糙集理论进行数据的约简,将简化的数据集作为BP神经网络输入数据,可简化BP神经网络的结构,同时缩短了样本训练时间,有效提高了BP神经网络分类正确率。在Matlab上进行仿真的结果表明,所提出的入侵检测算法,训练样本时间更短,入侵识别率和检测率却有了较以前的传统算法更高的准备率。

Intrusion detection system is of relatively higher false report rate and higher false alarm rate,and the defects of large data intrusion detection dimension,high redundancy and other exist.This paper proposes a new intrusion detection algorithm based on rough set theory and BP neural network.First,rough set theory is introduced based on domain concepts,thus reducing the loss of information.The use of rough set theory simplifies the input data set and the structure of BP neural network,reduces the training time,and effectively improves the classification accuracy of BP neural network.Matlab simulation results show that by usingthe proposed detection algorithm,the training time is shorter,and compared with the traditional network intrusion detection systems,the recognition rate and detection rate are better.