基于贝叶斯决策的交互式网络恶意入侵主动防御模型构建

Construction of an Interactive Network Malicious Intrusion Active Defense Model Based on Bayesian Decision Theory

为降低网络的恶意入侵风险,提出基于贝叶斯决策的交互式网络恶意入侵主动防御模型。采用K-聚类算法识别交互式网络中恶意入侵跳频数据,构建交互式网络恶意入侵节点分布模型;采用基于能量熵增量频域互相关系系数的敏感IMF分量选取算法,保留有效的恶意入侵特征分量。利用贝叶斯决策理念,构建恶意入侵防御模型,最终结果显示:该方法的抗干扰系数和冗余度结果分别在0.10和0.22以下;能够准确分类识别交互式网络中恶意入侵跳频数据;特征分量判定精度均在0.946以上;交互式网络的安全系数均在0.936;网络威胁等级均在2级以下,有效提升了网络的安全性。

In order to reduce network malicious intrusion risk,an interactive network malicious intrusion active defense model based on Bayesian Decision Theory is proposed.It uses K-means Clustering Algorithm to identify malicious intrusion frequency hopping data in interactive networks,constructs a distribution model of malicious intrusion nodes in interactive networks,and adopts a sensitive IMF Component Selection Algorithm based on the energy entropy increment frequency domain correlation coefficient to preserve effective malicious intrusion feature components.It uses Bayesian Decision Theory to construct a malicious intrusion defense model,and the final results show that the anti-interference coefficient and redundancy results of this method are below 0.10 and 0.22,respectively.It can accurately classify and identify malicious intrusion frequency hopping data in interactive networks,and the accuracy of feature component determination is above 0.946.The security factors of interactive networks are all 0.936.The network threat levels are all below level 2,effectively improving the security of the network.