Security Analysis of Cyber-Physical Systems Using Reinforcement Learning

Ibrahim, Mariam; Elhafiz, Ruba

doi:10.3390/s23031634

Cited by 9 publications

(6 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This helps in identifying the anomalies in the CPS. Reinforcement learning algorithms [42,43] can be used in CPS security and privacy to develop autonomous decision-making systems that can respond to changing environments and emerging threats. Reinforcement learning algorithms learn from feedback and reinforcement signals generated by the environment to adapt and improve their decision-making over time.…”

Section: Ai Techniques Used To Address the Security And Privacy Issuesmentioning

confidence: 99%

A Taxonomy of AI Techniques for Security and Privacy in Cyber–Physical Systems

Bandi

2024

JCCE

View full text Add to dashboard Cite

This research paper addresses the concerns related to security and privacy in cyber-physical systems (CPS) and explores the role of artificial intelligence (AI) in addressing these concerns. This paper presents a comprehensive classification of various security and privacy threats in CPS, providing an organized overview of potential risks, economic loss, and enabling effective risk assessment. This paper highlights how AI can help address the security and privacy concerns in CPS by presenting a detailed flow chart that illustrates the stepby-step process of using AI and machine learning (ML) techniques to detect security and privacy issues. This integrated approach serves as a guide for designing ML-based secure CPS, enabling proactive defense mechanisms and improving incident response and recovery. Furthermore, the research explores the various AI techniques that can be employed to address security and privacy concerns in CPS. A taxonomy of ML techniques specifically relevant to security and privacy issues is provided, offering insights into the potential applications of these techniques. In conclusion, this research emphasizes the significance of addressing security and privacy concerns in CPS and highlights the role of AI in tackling these challenges.

show abstract

Section: Ai Techniques Used To Address the Security And Privacy Issuesmentioning

confidence: 99%

A Taxonomy of AI Techniques for Security and Privacy in Cyber–Physical Systems

Bandi

2024

JCCE

View full text Add to dashboard Cite

show abstract

“…Amidst the evolving landscape of cybersecurity threats, concerted efforts are being made to harness the potential of RL to fortify cyber defenses. Ibrahim et al delve into the realm of cyber-physical systems with a focus on smart grid security, employing an RLaugmented attack graph to pinpoint vulnerabilities through the SARSA RL technique [34]. Dutta et al advance this frontier with a data-driven DRL framework designed to develop proactive, context-sensitive defense mechanisms that can adapt on-the-fly to changing adversary tactics, all while minimizing operational disruptions [35].…”

Section: Related Workmentioning

confidence: 99%

Employing Deep Reinforcement Learning to Cyber-Attack Simulation for Enhancing Cybersecurity

Oh,

Kim,

Nah

et al. 2024

Electronics

View full text Add to dashboard Cite

In the current landscape where cybersecurity threats are escalating in complexity and frequency, traditional defense mechanisms like rule-based firewalls and signature-based detection are proving inadequate. The dynamism and sophistication of modern cyber-attacks necessitate advanced solutions that can evolve and adapt in real-time. Enter the field of deep reinforcement learning (DRL), a branch of artificial intelligence that has been effectively tackling complex decision-making problems across various domains, including cybersecurity. In this study, we advance the field by implementing a DRL framework to simulate cyber-attacks, drawing on authentic scenarios to enhance the realism and applicability of the simulations. By meticulously adapting DRL algorithms to the nuanced requirements of cybersecurity contexts—such as custom reward structures and actions, adversarial training, and dynamic environments—we provide a tailored approach that significantly improves upon traditional methods. Our research undertakes a thorough comparative analysis of three sophisticated DRL algorithms—deep Q-network (DQN), actor–critic, and proximal policy optimization (PPO)—against the traditional RL algorithm Q-learning, within a controlled simulation environment reflective of real-world cyber threats. The findings are striking: the actor–critic algorithm not only outperformed its counterparts with a success rate of 0.78 but also demonstrated superior efficiency, requiring the fewest iterations (171) to complete an episode and achieving the highest average reward of 4.8. In comparison, DQN, PPO, and Q-learning lagged slightly behind. These results underscore the critical impact of selecting the most fitting algorithm for cybersecurity simulations, as the right choice leads to more effective learning and defense strategies. The impressive performance of the actor–critic algorithm in this study marks a significant stride towards the development of adaptive, intelligent cybersecurity systems capable of countering the increasingly sophisticated landscape of cyber threats. Our study not only contributes a robust model for simulating cyber threats but also provides a scalable framework that can be adapted to various cybersecurity challenges.

show abstract

“…The SARSA-based reward-extended attack graph algorithm was initially applied in our previous work [13] to identify the attacker's best course of action in a smart grid application. The availability of an attack graph created with knowledge of the resources, system design, connections, vulnerabilities, and potential threats is a determining factor.…”

Section: Related Workmentioning

confidence: 99%

Security Assessment of Industrial Control System Applying Reinforcement Learning

Ibrahim,

Elhafiz

2024

Processes

Self Cite

View full text Add to dashboard Cite

Industrial control systems are often used to assist and manage an industrial operation. These systems’ weaknesses in the various hierarchical structures of the system components and communication backbones make them vulnerable to cyberattacks that jeopardize their security. In this paper, the security of these systems is studied by employing a reinforcement learning extended attack graph to efficiently reveal the subsystems’ flaws. Specifically, an attack graph that mimics the environment is constructed for the system using the state–action–reward–state–action technique, in which the agent is regarded as the attacker. Attackers may cause the greatest amount of system damage with the fewest possible actions if they have the highest cumulative reward. The worst-case assault scheme with a total reward of 42.9 was successfully shown in the results, and the most badly affected subsystems were recognized.

show abstract

Security Analysis of Cyber-Physical Systems Using Reinforcement Learning

Cited by 9 publications

References 28 publications

A Taxonomy of AI Techniques for Security and Privacy in Cyber–Physical Systems

A Taxonomy of AI Techniques for Security and Privacy in Cyber–Physical Systems

Employing Deep Reinforcement Learning to Cyber-Attack Simulation for Enhancing Cybersecurity

Security Assessment of Industrial Control System Applying Reinforcement Learning

Contact Info

Product

Resources

About