Anomaly Detection of Industrial Control Systems Based on Transfer Learning

作     者：Weiping Wang Zhaorong Wang Zhanfan Zhou Haixia Deng Weiliang Zhao Chunyang Wang Yongzhen Guo Weiping Wang;Zhaorong Wang;Zhanfan Zhou;Haixia Deng;Weiliang Zhao;Chunyang Wang;Yongzhen Guo

作者机构：School of Computer and Communication Engineeringthe Beijing Key Laboratory of Knowledge Engineering for Materials Scienceand the Institute of Artificial IntelligenceUniversity of Science and Technology BeijingBeijing 100083China Shunde Graduate SchoolUniversity of Science and Technology BeijingGuangzhou 528399China School of Automation and Electrical EngineeringUniversity of Science and Technology BeijingBeijing 100083China School of Mechanical EngineeringUniversity of Science and Technology BeijingBeijing 100083China Donlinks School of Economics and ManagementUniversity of Science and Technology Beijing Beijing 100083China School of AutomationBeijing Institute of TechnologyBeijing 100081 China Software Testing CenterBeijing 100048China. 

出 版 物：《Tsinghua Science and Technology》 (清华大学学报（自然科学版（英文版）)

年 卷 期：2021年第26卷第6期

页      面：821-832页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0802[工学-机械工程] 0811[工学-控制科学与工程] 080201[工学-机械制造及其自动化] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported in part by 2018 industrial Internet innovation and development project“Construction of Industrial Internet Security Standard System and Test and Verification Environment” in part by the National Industrial Internet Security Public Service Platform in part by the Fundamental Research Funds for the Central Universities(Nos.FRF-BD-19-012A and FRFTP-19-005A3) in part by the National Natural Science Foundation of China(Nos.81961138010,U1736117,and U1836106) in part by the Technological Innovation Foundation of Shunde Graduate School,University of Science and Technology Beijing(No.BK19BF006) 

主　　题：anomaly detection transfer learning deep learning Industrial Control System(ICS) 

摘      要：Industrial Control Systems(ICSs)are the lifeline of a ***,the anomaly detection of ICS traffic is an important *** paper proposes a model based on a deep residual Convolution Neural Network(CNN)to prevent gradient explosion or gradient disappearance and guarantee *** developed methodology addresses two limitations:most traditional machine learning methods can only detect known network attacks and deep learning algorithms require a long time to *** utilization of transfer learning under the modification of the existing residual CNN structure guarantees the detection of unknown ***-dimensional ICS flow data are converted into two-dimensional grayscale images to take full advantage of the features of *** show that the proposed method achieves a high score and solves the time problem associated with deep learning model *** model can give reliable predictions for unknown or differently distributed abnormal data through short-term ***,the proposed model ensures the safety of ICSs and verifies the feasibility of transfer learning for ICS anomaly detection.