关于有向复杂网络软件异常交互行为检测研究  

Research on Abnormal Interaction Detection of Directed Complex Network Software

作　　者：崔素丽[1] 董东[2] 仇宾[1] 孙曼曼[1] CUI Su-li;DONG Dong;QIU Bin;SUN Man-man(Nationalities School,Hebei Normal University,Shijiazhuang Hebei 0500001,China;Hebei Normal University,Shijiazhuang Hebei 0500002,China)

机构地区：[1]河北师范大学附属民族学院,河北石家庄0500001 [2]河北师范大学,河北石家庄0500002

出　　处：《计算机仿真》2022年第11期294-298,共5页Computer Simulation

基　　金：河北省自然科学基金面上项目(F2019205303)。

摘　　要：针对复杂网络中软件运行时产生的交互行为,提出关于有向复杂网络软件异常交互行为检测方法。首先对三种类型的连边进行量化,依据有向路径的起始节点将网络路径分为四种类型,通过衰减参数对路径连接强度的相似性指标进行计算,完成对不同路径的区分。然后采用聚类方法创建候选状态,将信息设置到状态转移模型中,用状态划分算法将状态转移模型划分成多个二分图系统,并计算出各个子系统的网络最大流,标识出软件具体工作所产生的行为。最后构建描述异常行为交互过程的模型,将软件的执行过程视为组件的消息传输、组件关联等函数的调用,增强模型的准确性与完整性。实验结果表明,方法不仅能够准确地检测出软件的异常交互行为,还能准确地计算出各种参数的变化规律,对提高网络软件异常交互具有重要的意义。Aiming at the interaction behavior generated by software running in complex networks,a method for detecting abnormal interaction behavior of directed complex network software is proposed.Firstly,the three types of edge connection were quantified.According to the starting node of the directed path,the network path was divided into four types.The similarity index of the path connection strength was calculated through the attenuation parameters to distinguish different paths.Then,the candidate states were created by the clustering method,the information was set into the state transition model,the state transition model was divided into multiple bipartite graph systems by the state partition algorithm,the network maximum flow of each subsystem was calculated,and the behavior generated by the specific work of the software was identified.Finally,a model describing the interaction process of abnormal behavior was constructed,and the software execution process was regarded as the call of component message transmission,component association and other functions,so as to enhance the accuracy and integrity of the model.The experimental results show that this method can not only accurately detect the abnormal interaction behavior of software,but also accurately calculate the variation law of various parameters,which is of great significance to improve the abnormal interaction of network software.

关 键 词：衰减参数 聚类方法 转移模型 异常行为 交互过程 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]