基于大数据聚类的网络入侵检测技术研究  

Research on Network Intrusion Detection Technology Based on Big Data Clustering

作　　者：郑美容[1] ZHENG Mei-rong(Fujian Chuanzheng Communications College,Fuzhou 350000,China)

机构地区：[1]福建船政交通职业学院,福建福州350000

出　　处：《通化师范学院学报》2020年第12期65-71,共7页Journal of Tonghua Normal University

基　　金：2018年中青年教师教育科研项目(JZ180358).

摘　　要：针对网络入侵高维数据间差异显著,导致网络入侵检测存在所需迭代次数多、检测耗时长的问题,提出基于大数据聚类的网络入侵检测方法.结合基于大数据的网络数据预处理,将网络数据归一化、标准化;采用模糊C均值聚类算法,建立最大隶属原则,判断网络数据样本点是否异常,实现网络入侵检测.实验结果显示:所提方法在低维、高维数据空间中,对网络数据均可实现高精度、短时间的多种类入侵并行检测;且在检测网络入侵时,对网络数据的完整性存在较好保护.Due to the significant difference between high-dimensional data of network intrusion,network intrusion detection needs more iterations and takes longer detection time.This paper proposes a network intrusion detection method based on big data clustering.A network intrusion detection method based on big data clustering is proposed.Combined with the network data preprocessing based on big data,the net⁃work data is normalized and standardized.The fuzzy c-means clustering algorithm is used to establish the maximum membership principle and judge whether the network data sample point is abnormal,so as to re⁃alize the network intrusion detection.The experimental results show that the proposed method can realize high-precision and short-time multi-type intrusion parallel detection of network data in both low-dimen⁃sional and high-dimensional data Spaces.And when detecting network intrusion,the integrity of network data is well protected.

关 键 词：大数据 聚类 网络 入侵检测 数据预处理 模糊C均值 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]