基于逻辑推理的网络攻击想定生成系统Automated Network Attack.PDF

国防基础科研资助基金项目(项目名称和编号不公开)；北京教育委员会共建项目，北京市教育委员会， 基金编号：JD100060517 基于逻辑推理的网络攻击想定生成系统 1 1, 2 1 冯锦华 王海泉 孙芸芸 1.北京航空航天大学网络技术北京市重点实验室 北京 100083) (2.北京航空航天大学软件学院 北京 100083) 摘要：本文研究并实现基于逻辑推理的网络攻击想定自动生成系统，用于网络安全模拟演练。采用 Datalog 描述与安全相关的网络配置信息和攻击规则，使用XSB 推理引擎构造实现攻击目的的攻击路径 图；遍历攻击路径图， 到所有可能的攻击路径；根据攻击路径生成攻击想定脚本，由仿真平台解释并 执行相应的攻击操作序列。给出原型系统的设计与实现，及其实验结果。 关键字：逻辑推理 攻击规则 攻击路径 攻击路径图 攻击想定 中图分类号：TP393 文献标识码：A Automated Network Attack Scenario Generation System Based on Logical Reasoning Abstract: In this paper, Automated Network Attack Scenario Generation System is discussed, which has been implemented and will be used for network security exercise based on simulated network system. The target network s topology and configurations and attack rules are described in Datalog, which is a logical programming language. Using the powerful logical reasoning engine, XSB, An Attack Path Graph (APG) is constructed by analyzing proof records generated by XSB. All possible attack paths are generated through traversing APG and eliminating redundant attack operations existing in APG. According with the generated attack paths, Attack Scenario Scripts are generated and can be interpreted by simulation system to conduct attack operations. A prototype system has been implemented and experiment results are shown in this paper. Key words: Logical Reasoning, Attack Rule, Attack Path, Attack Path Graph (APG), Attack Scenario 1 引言 基于仿真网络的安全模拟演练是企业、单位或其它组织进行信息保障训练的一种重要手段。在国外