基于隐马尔可夫模型的网络安全态势预测方法.PDFVIP

第 46 卷第 10 期 中南大学学报( 自然科学版) Vol.46 No.10 2015 年 10 月 Journal of Central South University (Science and Technology) Oct. 2015 DOI: 10.11817/j.issn.1672-7207.2015.10.019 基于隐马尔可夫模型的网络安全态势预测方法 文志诚1,2，陈志刚1 (1. 中南大学 信息科学与工程学院，湖南 长沙，410083； 2. 湖南工业大学 计算机与通信学院，湖南 株洲，412007) 摘要：为了给网络管理员制定决策和防御措施提供可靠的依据，通过考察网络安全态势变化特点，提出构建隐马 尔可夫预测模型。利用时间序列分析方法刻画不同时刻安全态势的前后依赖关系，当安全态势处于亚状态或偏离 正常状态时，采用安全态势预测机制，分析其变化规律，预测系统的安全态势变化趋势。最后利用仿真数据，对 所提出的网络安全态势预测算法进行验证。访真结果验证了该方法的正确性。 关键词：网络安全态势；隐马尔可夫；态势预测；参数学习；预测模型 中图分类号：TP311 文献标志码：A 文章编号：1672−7207(2015)10−3689−07 Network security situation prediction method based on hidden Markov model WEN Zhicheng1, 2, CHEN Zhigang1 (1. School of Information Science and Engineering, Central South University, Changsha 410083, China; 2. School of Computer and Communication, Hunan University of Technology, Zhuzhou 412007, China) Abstract: In order to help network administrators make correct decisions and take effective defense measures, a hidden Markov prediction model was put forward. The characteristics of network security situation changes were investigated, the time series analysis method was used to describe the dependent relationship between the former and the latter’s security situations in different time. When the security situation was deviated from its normal state, the change law was analyzed, and system change trend and development direction of security situation in the future were predicted by the prediction model. Finally, the network security situation prediction algorithm was verified usi