Skip to content

This repository presents research on a novel composite framework for enhancing cyber resilience in Cyber Physical Systems (CPS). The framework integrates Dynamic Access Control, utilizing Authorizing Workflow and Task-Role-Based Access Control (AW-TRBAC), with Intrusion Detection Systems (IDSs) powered by Semantic Variational Autoencoders (SVAEs).

Notifications You must be signed in to change notification settings

LaBackDoor/svae-based-dynamic-threat-detection

Repository files navigation

Enhancing Cyber Resilience in Cyber Physical Systems

Abstract

This research introduces a composite framework designed to enhance cyber resilience in Cyber Physical Systems (CPS). The framework integrates Dynamic Access Control and Intrusion Detection Systems (IDSs) to improve CPS security and resilience against advanced persistent threats.

Key Components

Authorizing Workflow and Task-Role-Based Access Control (AW-TRBAC)

  • Redefines traditional access control
  • Incorporates dynamic Segregation of Duties (SoD)
  • Enables real-time decision-making
  • Offers context-aware control over system access

Semantic Variational Autoencoders (SVAEs)

  • Enhances IDS capabilities
  • Distinguishes between normal and anomalous system behaviors

Framework Characteristics

  • Multi-layered approach to resilience
  • Designed to prevent, withstand, and adapt to cyber threats
  • Dynamic nature allows real-time adjustments in access control policies
  • Probabilistic learning capacity for in-depth analysis of CPS access flows

Validation

  • Simulations conducted in water treatment CPS scenarios
  • Demonstrated effectiveness in improving security posture
  • Minimal false positives in intrusion detection
  • Enhanced access control

Contributions

This research presents a security solution that addresses the unique challenges of Cyber-Physical Systems, ultimately bolstering cyber resilience in this critical domain.

Citation

If you use the code or concepts from this research in your work, please cite our paper. 📜

Plain Text

A. Orojo, E. El-Mahmoud, S. Hutton, W. Elumelu, and M. Donahoo. (2025). "A Unified Framework Incorporating AW-TRBAC and Semantic Variational Autoencoders for Dynamic Threat Detection and Access Control." International Conference on Artificial Intelligence 2025.

BibTeX

@inproceedings{orojo2025unified,
  title={A Unified Framework Incorporating AW-TRBAC and Semantic Variational Autoencoders for Dynamic Threat Detection and Access Control},
  author={Orojo, A and El-Mahmoud, E and Hutton, S and Elumelu, W and Donahoo, M},
  booktitle={International Conference on Artificial Intelligence 2025},
  year={2025}
}

About

This repository presents research on a novel composite framework for enhancing cyber resilience in Cyber Physical Systems (CPS). The framework integrates Dynamic Access Control, utilizing Authorizing Workflow and Task-Role-Based Access Control (AW-TRBAC), with Intrusion Detection Systems (IDSs) powered by Semantic Variational Autoencoders (SVAEs).

Topics

Resources

Stars

Watchers

Forks