Industry 4.0 represents the fourth phase of industry and manufacturing revolution, unique in that it provides Internet-connected smart systems, including automated factories, organizations, development on demand, and ‘just-in-time’ development. Industry 4.0 includes the integration of cyber-physical systems (CPSs), Internet of Things (IoT), cloud and fog computing paradigms for developing smart systems, smart homes, and smart cities. Given Industry 4.0 is comprised sensor fields, actuators, fog and cloud processing paradigms, and network systems, designing a secure architecture faces two major challenges: handling heterogeneous sources at scale and maintaining security over a large, disparate, data-driven system that interacts with the physical environment. This paper addresses these challenges by proposing a new threat intelligence scheme that models the dynamic interactions of industry 4.0 components including physical and network systems. The scheme consists of two components: a smart management module and a threat intelligence module. The smart data management module handles heterogeneous data sources, one of the foundational requirements for interacting with an Industry 4.0 system. This includes data to and from sensors, actuators, in addition to other forms of network traffic. The proposed threat intelligence technique is designed based on beta mixture-hidden Markov models (MHMMs) for discovering anomalous activities against both physical and network systems. The scheme is evaluated on two well-known datasets: the CPS dataset of sensors and actuators and the UNSW-NB15 dataset of network traffic. The results reveal that the proposed technique outperforms five peer mechanisms, suggesting its effectiveness as a viable deployment methodology in real-Industry 4.0 systems.
A New Threat Intelligence Scheme for Safeguarding Industry 4.0 Systems
Nour Moustafa,Erwin Adi,B. Turnbull,Jiankun Hu
Published 2018 in IEEE Access
ABSTRACT
PUBLICATION RECORD
- Publication year
2018
- Venue
IEEE Access
- Publication date
2018-06-07
- Fields of study
Computer Science, Engineering, Environmental Science
- Identifiers
- External record
- Source metadata
Semantic Scholar
CITATION MAP
EXTRACTION MAP
CLAIMS
CONCEPTS
- anomalous activities
Suspicious behavior patterns in the monitored systems that the model is designed to identify.
Aliases: anomalies
- beta mixture-hidden markov models
A probabilistic modeling approach used here for identifying anomalous activity patterns over time.
Aliases: MHMMs
- cps dataset
A benchmark dataset containing sensor and actuator data from cyber-physical systems.
Aliases: the CPS dataset
- five peer mechanisms
Five comparison methods used as baselines for evaluating the proposed approach.
- industry 4.0 systems
Internet-connected industrial environments that integrate cyber-physical, sensing, and networking components.
Aliases: Industry 4.0
- physical and network systems
The combined physical-process and networked communication layers considered by the detection scheme.
Aliases: physical system and network system
- smart management module
The component that handles heterogeneous data sources and data flow within the proposed scheme.
- threat intelligence module
The component that performs threat detection and anomaly discovery in the proposed scheme.
- threat intelligence scheme
An architecture for detecting and managing threats in Industry 4.0 environments by modeling interactions among system components.
- unsw-nb15 dataset
A benchmark network-traffic dataset used for evaluating intrusion-detection methods.
Aliases: the UNSW-NB15 dataset
REFERENCES
Showing 1-46 of 46 references · Page 1 of 1