Peer-to-Peer Federated Learning Based Anomaly Detection for Open Radio Access Networks

Open radio access network (O-RAN) has been recognized as a revolutionized architecture to support the multi-class wireless services required in fifth-generation (5G) and beyond 5G networks. The openness and the distributed nature of the O-RAN architecture have created new forms of threat surfaces than the conventional RAN architecture and require complex anomaly detection mechanisms. Moreover, with the introduction of RAN intelligent controllers (RICs), it is possible to utilize advanced Artificial Intelligence (AI)/ Machine Learning (ML) algorithms based on closed control loops to detect anomalies in a data-driven manner. In this paper, we particularly investigate the use of Federated Learning (FL) for anomaly detection in the O-RAN architecture, which can further preserve data privacy. We propose a peer-to-peer (P2P) FL-based anomaly detection mechanism for the O-RAN architecture and provide a comprehensive analysis of four variants of P2P FL techniques. Moreover, we simulate the proposed models using the UNSW-NB15 dataset.