AI and Digital Twin-based Multi-Operator Collaboration for 6G Use Cases

Sixth-generation (6G) networks must support immersive and mission-critical services that impose extreme and simultaneous requirements on data rate, latency, reliability, and connectivity. Meeting these stringent demands challenges the capabilities of traditional single-operator network operation, necessitating architectural enhancements. To address this, this paper investigates multi-operator collaboration as a scalable paradigm for 6G, enabled by the synergy of artificial intelligence (AI) and digital twins (DTs). We propose a novel, unified framework: per-operator DTs facilitate high-fidelity, predictive network modeling, while a federated multi-agent deep reinforcement learning scheme enables robust, privacy-preserving cross-operator optimization. A real-world case study, utilizing extensive measurement data from four major mobile network operators (MNOs) in the United Kingdom, demonstrates the framework’s efficacy. Results show that the proposed multi-MNO collaborative approach significantly reduces quality-of-service violations and improves achievable data rates compared to standalone deployments, particularly under the high-throughput demands characteristic of immersive applications. Finally, we conclude by discussing the key challenges and future research directions essential for realizing practical, AI-native 6G networks.