面向复杂网络的威胁度量及聚合方法  

Similarity measures and polymerization to identity threats in complex networks

作　　者：邓辉[1] 刘晖[1] 张宝峰[1] 毛军捷[1] 郭颖[1] 熊琦[1] 谢仕华[1] 

机构地区：[1]中国信息安全测评中心,北京100085

出　　处：《清华大学学报（自然科学版）》2016年第5期511-516,共6页Journal of Tsinghua University(Science and Technology)

基　　金：国家自然科学基金资助项目(61472448)

摘　　要：在复杂网络中,威胁模型结构庞大、行为复杂,不利于建模后的威胁分析。该文从实现的角度出发,针对一类利用C程序实现的威胁对象及威胁,在已有的威胁建模理论的基础上,基于代数系统理论提出威胁对象及威胁的代数化刻画框架。基于该框架,采用代数簇理论建立威胁行为相似度度量函数,通过矩阵理论及非线性约束求解理论进行函数求解,从而实现相似行为的代数化判定。最后,针对判定后的相似行为,基于并发系统等价关系构建威胁行为聚合规则,实现威胁模型优化,减少威胁分析复杂度优化。The huge structures and the complex behavior of threat models in complex networks are given too much computing effort for threat analyse. This paper presents an algebraic framework for threat modeling using algebraic theory to describe the object and its threats which are all implemented in a C program. An algebraic function measures the similarities among different threats and then expands the analysis using matrixes or nonlinear constraint theory. Finally, an equivalence relation for the concurrent theoretical is used to established a threat polymerization rule for similar threats to optimize the threat model and reduce the threat analysis complexity.

关 键 词：威胁建模 相似度度量 威胁聚合 威胁分析 

分 类 号：TP301.2[自动化与计算机技术—计算机系统结构]