摘要
影响多节点网络安全态势的因素有很多,传统方法未考虑影响因素,导致预测精度低。为此,构建一种新的改进G-K算法的多节点网络安全态势预测模型。通过灰熵均衡关联度实现关联分析,通过关联分析法对影响多节点网络安全态势的因素进行研究,按照关联程度大小获取重要因素。依据多节点网络日志信息,从主机层、服务层和系统层对多节点网络安全态势值进行计算。分析改进G-K算法预测基本思想:通过多节点网络安全态势重要影响因子值与安全态势值计算状态向量,利用当前状态矢量观测向量对多节点网络安全态势在下一段时间的值进行描述。在此基础上,构建改进G-K算法的多节点网络安全态势预测模型。实验结果表明,采用构建的模型对多节点网络安全态势进行预测,有很高的预测精度,对突变态势也可有效预测。
There are many factors affecting the security situation of multi node network.The traditional methods do not consider the influence factors and lead to low prediction accuracy.To this end,a new multi node network security situation prediction model based on improved G-K algorithm is constructed.Through grey entropy,the correlation analysis is implemented,and the factors that influence the multi node network security situation are studied through correlation analysis.According to the multi node network log information,the multi node network security situation value is calculated from the host layer,the service layer and the system layer.The basic idea of the improved G-K algorithm is analyzed.The state vector is calculated through the multi node network security situation important influence factor value and the security situation value,and the current state vector observation vector is used to describe the value of the multi node network security situation in the next period.On this basis,a multi node network security situation prediction model based on improved G-K algorithm is constructed.The experimental results show that the construction model is used to predict the security situation of the multi node network,which has high prediction accuracy and can be effectively predicted for the sudden abnormal potential.
作者
周新卫
李小玲
ZHOU Xin-wei;LI Xiao-ling(Department of information engineering,Nanchang University Gongqing college,Gongqingcheng 332020,China)
出处
《科学技术与工程》
北大核心
2018年第25期72-77,共6页
Science Technology and Engineering
基金
2015年江西省教育厅教改项目(JXJG-15-31-1)资助
关键词
改进G-K算法
多节点
网络
安全态势
预测模型
improved G-K algorithm
multi node
network
security situation
prediction model
