基于M-ATS的动态网络伪装攻击挖掘算法  

Dynamic Network Camouflage Attack Mining Algorithm Based on M-ATS

作　　者：苏文明 SU Wenming(School of Information and Finance,Xuancheng Vocational and Technical College,Xuancheng Anhui 242000)

机构地区：[1]宣城职业技术学院信息与财经学院,安徽宣城242000

出　　处：《湖北理工学院学报》2023年第2期21-24,58,共5页Journal of Hubei Polytechnic University

摘　　要：为了解决动态网络数据复杂性、伪装攻击难以识别的问题,提出了基于M-ATS的动态网络伪装攻击挖掘算法。采用转化动态网络中字符型数据的方式,去除数值型数据冗余特征,归一化预处理动态网络数据,并采用k-means聚类算法对动态网络数据进行增益,获取伪装攻击的有效节点数据,利用M-ATS模型对动态网络节点的数据进行伪装攻击信息挖掘,通过迭代优化的方式确定M-ATS模型的全局最优参数,结合动态网络各个节点相关回报值的波动性,获取具体伪装攻击信息特征,实现对伪装攻击的准确挖掘。仿真测试结果表明,算法对动态网络安全态势值的计算结果与伪装攻击的结果具有较高的拟合度,应用效果良好,可以为动态网络伪装攻击挖掘提供参考。In order to solve the problem of the complexity of dynamic network data and difficulty in identifying camouflage attacks,this paper proposes a dynamic network camouflage attack mining algorithm based on M-ATS.The character-based data in the dynamic network is transformed to remove the redundant features of numerical data,normalize the pre-processed dynamic network data,and use the k-means clustering algorithm to gain the dynamic network data to obtain the effective node data of the masquerade attack.The M-ATS model is used to mine the data of the dynamic network nodes for the masquerade attack information,determine the M-ATS by iterative optimization.The global optimal parameters of the model are determined by iterative optimization,and the volatility of the relevant return values of each node of the dynamic network is combined to obtain the specific disguise attack information features and achieve accurate mining of the disguise attack.The simulation test results show that the algorithm has a high fit between the calculation results of the dynamic network security posture values and the results of the masquerade attack,and the application effect is good,providing a reference for the mining of dynamic network camouflage attacks.

关 键 词：动态网络 伪装攻击 归一化 K-MEANS聚类算法 回报值 自适应 

分 类 号：TP393.1[自动化与计算机技术—计算机应用技术]