基于攻击意图的CPS跨域攻击路径行为分析方法  

CPS Cross Domain Attack Path Behavior Analysis Method Based on Attack Intention

作　　者：何定坤 周纯杰[1] He Dingkun

机构地区：[1]华中科技大学人工智能与自动化学院,湖北武汉430074

出　　处：《工业控制计算机》2021年第6期10-12,共3页Industrial Control Computer

基　　金：云计算环境下工业信息物理系统主动入侵反应关键技术及应用研究(61873103)。

摘　　要：随着工业互联网、云计算、5G等新技术出现,工业控制系统已逐步从封闭隔离系统演进为开放交互系统,在智能化、网络化的同时面临着更多新的信息安全威胁。以攻击者攻击意图为基础,提出CPS跨域攻击路径行为分析框架,从攻击手段和攻击路径预测两个维度进行分析:通过节点与系统的双重隐马尔可夫模型分析下一步的攻击手段,通过元胞自动机的状态更新分析CPS攻击的跨域传播,二者的预测结果同时作为反馈验证攻击者的攻击意图。With the in-depth integration of industrial control system infrastructure network and intelligence,the increase in openness has led to an increasing degree of coupling and mutual influence between information systems and physical systems,making traditional industrial control systems face more new information security threats.This paper proposes a CPS cross-domain attack path behavior analysis framework based on the attacker's attack intentions,and analyzes the attack means and attack path prediction from two dimensionsthe next step of attack means is analyzed through the dual state hidden Markov model of the node and the system.Analyze the cross-domain propagation of CPS attacks through the state update of the cellular automata,and the prediction results of the two are used as feedback to verify the attacker's attack intention.

关 键 词：攻击路径行为分析 双重隐马尔可夫模型 元胞自动机 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]