A Novel Network Proactive Defense Model: Anomaly Driven Dynamic Cooperative Defense Model

2018 2nd International Conference on Artificial Intelligence Applications and Technologies

计算机科学

Lixun, Li^1,2 ; Bin, Zhang^1,2 ; Haoming, Wu^1,2 ; Yingbo, Li^1,2

Zhengzhou Information Science and Technology Institute, ZhengZhou, China^1

Henan Province Information Security Key Laboratory, ZhengZhou, China^2

关键词: Formal model;    Moving target defense;    Multi-step attacks;    Mutiagent systems;    Proactive defense;    System security;    Typical application;    Unknown networks;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/435/1/012064/pdf
DOI  :  10.1088/1757-899X/435/1/012064

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

To defend the more complex and stealthy network attacks, an anomaly driven dynamic cooperative defense model (ADCD) is proposed in this paper. ADCD takes anomalies as defense drive factors to identify network attack types and evaluate system security state. And according to the anomaly processing result, ADCD can combine network cooperative defense (NCD) with moving target defense (MTD) to defend known and unknown network attacks. Firstly, detail explanations of ADCD is given, including its elements, operations, relations and rules. Then, a formal analysis of ADCD is conducted based on finite state machine (FSM) to prove its security. Finally, A typical application scenario where ADCD is realized as a mutiagent system is displayed to verify the effectiveness of ADCD in defending complex and multi-step attacks.