基于云模型的入侵检测研究  

Study of Intrusion Detection Based on Cloud Model

作　　者：蒋建兵[1] 粱家荣 王龙[1] 

机构地区：[1]广西柳州市广西工学院,545006 [2]广西大学计电学院,530004

出　　处：《微计算机信息》2010年第3期102-103,共2页Control & Automation

摘　　要：由于网络行为的不确定性,使现有入侵检测系统几乎都存在高误报率和高漏报率的缺点。云模型是将模糊性和随机性有机结合进行不确定性推理的有效工具。本文利用云模型来处理网络实体行为的不确定性,提出了一种新的云入侵检测方法。该方法通过基于云知识库的云推理引擎进行不确定性推理,以对网络实体行为进行智能判断。模拟结果表明该方法能有效提高入侵检测效率。Because of the uncertainty of network behavior,the exiting intrusion detection systems almost have some flaws,such as the high false positive rate and high false negative rate. Cloud model is an effective tool of uncertainty reasoning in transforming between qualitative concepts and their quantitative expressions combined with fuzziness and randomness. Based on cloud model,a new cloud intrusion detection method（CIDM） was presented in this paper. Cloud reasoning generator cloud intelligently processes network behavior by means of uncertainty reasoning based on cloud knowledge database. The simulation results indicate CIDM can greatly improve the effectiveness of intrusion detection.

关 键 词：入侵检测 网络安全 不确定性推理 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]