RapidRANDefender: AI security for Open networks

Abstract

The Open Radio Access Network (O-RAN) architecture represents a keystone paradigm in establishing more flexible, versatile, and intelligent next-generation wireless communication networks. By adopting an open platform, O-RAN reduces total costs and lowers barriers for new vendors to enter the market. As a result, O-RAN is projected to account for up to 30% of worldwide RAN revenues by 2028\. While the integration of AI within the O-RAN architecture holds great promise for fostering innovative services, it also introduces significant security challenges. Protecting the logic behind the O-RAN Radio Intelligent Controller (RIC) is crucial, as any compromise could severely impact network performance and, in extreme cases, endanger human lives. This project addresses these security concerns by developing a novel strategy based on abusive adversarial agents operating within a digital twin replica of the underlying network. This approach enables low-cost, real-time threat detection without requiring direct access to the actual network or disrupting normal operations. Specifically, it allows for the interception and notification of all potential current and future vulnerabilities, enabling proactive cyber-attack prevention and safeguarding the underlying network infrastructure.

Lead Participant

Project Cost

Grant Offer

QUEEN MARY UNIVERSITY OF LONDON £18,985 £ 18,985
 

Participant

QUEEN'S UNIVERSITY OF BELFAST £16,000 £ 16,000

Publications

10 25 50