ARMOR: Advancing AI-Driven Security for 6G Open RAN Networks

News

ARMOR: Advancing AI-Driven Security for 6G Open RAN Networks

UCD NetsLab’s ARMOR (Adversarial Resistance and Model Optimization for Robustness in 6G Open Radio Access Networks) project has been successfully funded under the 6G-SANDBOX 3rd Open Call. This funding enables ARMOR to pioneer AI-driven security solutions for future 6G networks, focusing on identifying and mitigating adversarial threats targeting AI models deployed in O-RAN environments.

Advancing AI Security in Open and Programmable Networks

As AI becomes a fundamental component of 6G telecommunications, its vulnerability to adversarial attacks, data poisoning, and privacy breaches poses significant security challenges. ARMOR is among the first initiatives dedicated to tackling AI-specific threats in 6G O-RAN, ensuring robust, trustworthy, and resilient AI-driven network infrastructures.

The project introduces an AI security testing framework, integrating adversarial attack simulations, Explainable AI (XAI), and automated vulnerability reporting to enhance AI model robustness in network optimization, security enforcement, and service management.

Key Objectives of ARMOR

✅ Developing a Comprehensive Adversarial Testing Framework

✅ Identifying and Mitigating AI Model Vulnerabilities

✅ Ensuring Safe Deployment of 6G Networks

✅ Providing Actionable Security Insights for Network Operators

Testing in Leading 6G Experimental Platforms

ARMOR will be validated in real-world 6G testbeds, including:

UCD NetsLab’s 5G ORAN Test Network, analyzing AI-driven xApp security.
6G-SANDBOX remote test sites, enabling large-scale security evaluations across European research networks.

Strengthening 6G AI Security through Collaboration

“With the rise of AI-powered telecommunications, securing these models from adversarial threats is essential,” said Dr. Madhusanka Liyanage, Principal Investigator of ARMOR. “By leveraging 6G-SANDBOX funding, ARMOR will deliver an AI security framework that ensures resilience, privacy, and trustworthiness in future network deployments.”

The ARMOR project will be conducted by PhD students Chamara Sandeepa and Charuka Moremada.