CERBERUS: Towards Secure and Automated Multi-operator Management in B5G Through a Dynamic Policy-Based ZSM Framework

The evolving landscape of 5G and Beyond-5G (B5G) networks is characterized by high connectivity speeds, low latencies, and integration with diverse infrastructures such as smart cities and IoT. However, these advances are also exploited by increasingly sophisticated security threats, making it essential to evolve cybersecurity approaches. This involves not only automating security measures within each operator, but also enabling proactive and reactive orchestration mechanisms that enable cooperation and information sharing across multiple operators. Building upon earlier work on CERBERUS, this extended paper provides a more comprehensive description of the framework. Specifically, it details the architecture, implementation, and experimental validation of a collaborative, intelligent, multi-operator platform capable of offering real-time, autonomous responses to both existing and emerging threats. Leveraging zero-touch service management principles and Q-learning-based decision-making, the framework automates key processes in service, security, and network management. It supports real-time monitoring, mitigation, prevention, and coordinated response to cyber threats, with a strong focus on trust management and dynamic multi-operator security agreements. Results, based on the use case presented, show that the framework enables proactive orchestration (including dynamic VNF deployments) in less than 35 s, detects attacks in approximately 1 s, and applies SDN-based mitigations in under 2.5 s. Moreover, threat information sharing and monitoring reconfiguration across operators are achieved in less than 10 s, demonstrating its effectiveness in real-time, collaborative security management for 5G environments.