DEWC Services, in collaboration with the University of Adelaide and Defence Trailblazer, is leading an advanced research initiative into hybrid moving target defence (MTD) for container-based cloud environments. This project leverages reinforcement learning to develop an AI-driven cyber defence framework that adapts dynamically to evolving threats.
As cyber threats grow in sophistication, traditional static defence mechanisms are proving inadequate in modern cloud environments. MTD disrupts adversarial attacks by continuously modifying the attack surface, making it significantly harder for attackers to penetrate systems.
While MTD has been explored extensively in cyber security research, existing models struggle with scalability, cost-effectiveness, and real-world application. This initiative aims to overcome these limitations by applying reinforcement learning, allowing for autonomous, intelligent, and adaptive cyber defence mechanisms.
“The integration of reinforcement learning into Hybrid MTD represents a major step forward in cyber resilience,” said DEWC Services’ Dr Tim McKay. “By enabling an AI-driven approach, we can learn optimal MTD configurations, improving security posture and reducing vulnerabilities in cloud-based defence systems.”
The research focuses on developing a reinforcement learning framework to dynamically select and optimise configurations of MTD strategies based on real-time feedback from the environment. Key objectives include:
- Optimising hybrid MTD deployment to enhance security effectiveness while maintaining operational efficiency;
- Ensuring easy integration with traditional cyber security frameworks;
- Developing cost-effective solutions suitable for long-term defence and national security applications; and
- Validating hybrid MTD strategies through testing in high-threat, real-world scenarios.
“This research will allow us to explore various defence strategies in complex cyber threat scenarios,” said Associate Professor Claudia Szabo of the Lead Complex Systems Research Group at the University of Adelaide. “Our focus is on understanding how reinforcement learning can enhance adaptability and efficiency in cyber defence, ensuring Australia remains at the forefront of cyber security innovation.”
DEWC Services will present its research progress at the inaugural Defence Trailblazer Technical and Capability Conference on 18 March. This presentation will highlight shortcomings in existing MTD approaches and discuss how reinforcement learning-enabled hybrid MTD can enhance cyber resilience in container-based cloud environments.
DEWC’s participation in Defence Trailblazer is part of its commitment to workforce skilling programs for defence, recognising that a highly skilled workforce is essential for meeting evolving national security challenges.
“This project is a key example of how Defence Trailblazer is facilitating critical research that will directly benefit national security,” said Defence Trailblazer Executive Director Dr Sanjay Mazumdar. “By investing in cutting-edge AI-driven cyber defence, we are ensuring the Australian Defence Force and its partners have access to the most advanced security solutions available.”
As the research progresses, the team will develop a proof-of-concept prototype, demonstrating the practical deployment of hybrid MTD strategies in real-world defence applications. By leveraging AI-driven security, this project represents a paradigm shift in cyber defence, reinforcing Australia’s resilience against future cyber threats.
