稳态网络抗攻击频率准确预测方法仿真

Static Network Anti-attack Frequency Accurate Prediction Method Simulation

以解决当前网络抗攻击频率预测方法中存在的预测准确性差、能耗高的问题为目的,提出基于贝叶斯理论的稳态网络抗攻击频率预测方法。将网络攻击报警数据划分为以攻击报警的数据名称为特征、以攻击目的 IP地址为特征、以攻击时间间隔为特征的三个类别。分别计算两个攻击报警事件间名称的相似隶属函数值、攻击目的 IP地址的相似隶属函数值、攻击时间间隔相似隶属函数值,并获取两个攻击报警事件间的总相似隶属度函数值,完成攻击报警数据聚合。将攻击报警聚合代入攻击证据计算中,并将攻击证据当作节点,组建贝叶斯网络蓄意攻击图。利用攻击图对攻击者于攻击时的攻击路径以及攻击状态进行分析,通过贝叶斯理论计算稳态网络内部危险后验概率,将后验概率和能耗控制阈值引入网络抗攻击频率的预测中,实现网络抗攻击预测。实验表明,上述方法预测准确性系数最高为0. 98,所用能耗低。所提方法鲁棒性强,具有可实践性。

To solve the problem of poor prediction accuracy and high energy consumption in the current network anti-attack frequency prediction methods,a steady-state network anti-attack frequency prediction method based on Bayesian theory is proposed. The network attack alarm data is divided into three categories characterized by the data name of the attack alarm,the characteristics of the attack destination IP address,and the characteristics of the attack time interval. Calculate the similar membership function values of the names of the two attack alarm events,the similar membership function values of the attack destination IP address,and the similar membership function values of the attack time interval,and obtain the total similar membership function values between the two attack alarm events. Attack alarm data aggregation. The attack alarm aggregation is substituted into the attack evidence calculation,and the attack evidence is taken as a node to form a Bayesian network deliberate attack graph. Using the attack graph to analyze the attack path and attack state of the attacker,calculate the risk posteriori probability inside the steady network through Bayesian theory,and introduce the posterior probability and energy consumption control threshold into the prediction of the network anti-attack frequency. To achieve network anti-attack predictions. Experiments show that the method has a prediction accuracy coefficient of up to 0. 98 and low energy consumption. The proposed method is robust and practical.