Fault Detection, Diagnosis, and Prediction for IP-based Industrial Control Networks

DoctorMission-critical Industrial Control Networks (ICN) support secure and reliable communications of devices in process control or manufacturing environments. ICN has been heavily dependent on the proprietary protocols from various vendor-specific technologies. Recently, however, many of these proprietary protocols are being migrated to IP networks in order to consolidate various different types of networks into a single common network. This process has been undertaken to simplify the network operation, administration and maintenance and to reduce operational expenses and capital expenditures. Despite the wide deployment of ICN, most operators have very little knowledge on how to manage their ICN reliably and securely. This is mainly due to the operators’ unfamiliarity with the various faults that occur on ICN and their defensive network maintenance strategy. The current process of detecting and diagnosing faults is mostly manual and thus the operators generally detect problems only after noticeable malfunctioning has occurred. In addition, the existing IP diagnosis techniques have not been able to fully handle fault symptoms and mainly focus on network diagnostics rather than process or device diagnostics. This thesis presents that the absence of advanced fault diagnosis techniques leads to the development of new methodologies which are suitable for ICN. We describe unique traffic characteristics and categorize the faults of ICN. We also propose novel fault diagnosis, prediction, and adaptive decision methodologies, and verify them with real-world ICN data from POSCO. Finally, this thesis proposes fault diagnosis system architecture for ICN. Our experience in developing the fault diagnosis system provides a firm guideline to understand the fault management mechanisms in a large ICN. Overall, this thesis presents a complete cycle of handling faults in IP networks within the scope of ICN: Monitoring, analysis, and prediction