Federated Network Intelligence Orchestration for scalable and automated FL-based anomaly detection in B5G Networks
Pablo Fernández Saura, José Manuel Bernabé Murcia, Alejandro Molina Zarca, Jorge Bernal Bernabé, Antonio Fernando Skarmeta Gomez · 2023
Network Intelligence management in Beyond 5G networks embraces the exciting challenge of addressing scalability, dynamicity, interoperability, privacy, and security concerns. These are essential steps towards achieving the realization of truly ubiquitous AI-based analytics, empowering seamless integration across the entire Continuum (Edge, Fog, Core, Cloud). To address these challenges, this paper presents a model-driven Federated learning approach for managing and Orchestrating the Network intelligence needed to detect and prevent cyber-attacks. The system has been integrated into a B5G Security Framework, leveraging the multi-domain and multi-tenant Orchestrator thereby endowing the Network intelligence with key features for automated and scalable deployment of FL-agents and AI-based anomaly detectors, strengthening the reaction capabilities to counter cyber-attacks. The presented FL-based model-driven system allows interoperability and extensibility in the management of the FL system.