网络级联抗毁攻击信息层次化分类仿真研究

Simulation Research on Hierarchical Classification of Network Cascade Anti-Destruction Attack Information

网络技术加速了社会信息化进程。但是网络安全形势日趋严峻,网络攻击技术正在不断地向隐蔽、高效和智能化方向发展。网络级联抗毁攻击信息层次化分类是骨干网安全监测条件下海量报警信息处理的迫切需要。针对平面分类方法对网络级联抗毁攻击信息分类时出现的精度不高的现象,设计一种基于程序树的层次化分类方法。采用分词管理的方式,将网络级联抗毁性攻击信息作为一种特征项考虑,提取攻击信息的节点特征;将攻击信息的权重计算过程看做是一个节点信息的频度权衡过程,计算攻击信息节点权重值,利用程序树的结构特征,按照根、茎、叶的分类方式,实现网络级联抗毁攻击信息的层次分类。仿真结果表明,层次分类方法比平面分类方法的信息分类精度提高了15.16%。说明基于程序树的层次化分类方法具备有效性,能够实现网络级联抗毁攻击信息的准确分类。

In recently years,network technology has accelerated the process of social informatization,but the network security situation is becoming more and more serious.The network attack technology is developing to the efficiency and intelligence.It is an urgent need for massive alarm information processing to classify the network cascade invulnerability attack information in a hierarchical manner.Due to low accuracy of planar classification method,this article presented a hierarchical classification method based on program tree.Firstly,we used the word segmentation management method to consider the network cascading invulnerability attack information as a feature item,so as to extract the node characteristics of attack information.Secondly,we considered the weight calculation of attack information as a frequency weighing process of node information,and thus to calculate the weight value of attack information node.Based on the classification of roots,stems,and leaves,we used the structural characteristics of program tree to achieve the hierarchical classification of network cascading invulnerability attack information.Simulation results show that the classification accuracy of hierarchical classification method is 15.16% higher than that of the planar classification method,so the hierarchical classification method based on program tree can accurately classify the network cascading invulnerability attack information.