Asset-Aware and Resilient Trust Management Framework for Industrial IoT Edge Networks

Area of Science:

• Computer Science
• Network Security
• Industrial Internet of Things (IIoT)

Background:

• Trust evaluation in IIoT edge networks is complex, requiring consideration of device behavior and asset criticality.
• Existing uniform scoring models struggle with sophisticated attacks like semantic attacks and whitewashing, and impose high processing demands on edge devices.

Purpose of the Study:

• To present the Asset-Aware Resilient Trust (ART) framework for robust and efficient trust evaluation in IIoT edge networks.
• To address limitations of uniform scoring by separating behavioral credibility from asset criticality.

Main Methods:

• ART employs a dual-plane architecture to distinguish dynamic behavioral credibility from physical asset criticality.
• Cross-layer evidence is gathered from communication, identity, physical, and semantic interactions.
• A Fuzzy Triggered-Entropy Weight Method (Fuzzy T-EWM) adaptively weights evidence based on observed fluctuations, and trust scores are updated using a Fast-Drop Slow-Rise rule with a tolerance margin.

Main Results:

• ART effectively detects stealthy False Data Injection attacks and mitigates whitewashing behavior.
• The framework significantly reduces computational overhead by 76.4% compared to the Standard EWM baseline.
• The credibility-weighted aggregation mechanism prevents collusive manipulation during cold-start evaluations.

Conclusions:

• The ART framework provides differentiated trust regulation for IIoT edge networks, enhancing security and efficiency.
• ART's dual-plane architecture and adaptive weighting method offer a more resilient and computationally efficient approach to trust management.