Fortifying Open RAN Security with Zero Trust Architecture and Transformers

Open Radio Access Networks (Open RAN) have gained significant traction in modern telecommunications due to their cost-efficiency and scalability. However, their open architecture introduces critical security challenges. To fortify Open RAN security, this paper proposes a novel approach that combines the principles of Zero Trust Architecture (ZTA) with the use of transformer-based models. ZTA operates on the premise that all network entities are untrusted until authenticated, providing a robust security foundation. By integrating transformers, renowned for their efficiency in processing unstructured data, we enable real-time analysis of time series data within specified windows, enhancing anomaly detection. We leverage a dataset constructed from a functional 5G test network to evaluate the effectiveness of our transformer-based approach. Our findings demonstrate the superior performance of our approach compared to other benchmark methods.