AI-driven Closed-loop Automation in 5G and beyond Mobile Networks

Abstract

The 5th Generation (5G) mobile networks support a wide range of services that impose diverse and stringent QoS requirements. This will be further exacerbated with the evolution towards 6th Generation mobile networks. Inevitably, 5G and beyond mobile networks must provide stricter, differentiated QoS guarantees to meet the increasing demands of future applications, which cannot be satisfied with traditional human-in-the-loop service orchestration and network management approaches. In this paper, we lay out our vision for closed-loop service orchestration and network management of 5G and beyond mobile networks. We extend the MAPE (i.e., monitor, analyze, plan, and execute) control loop to facilitate closed-loop automation, and discuss the quintessential role of Artificial Intelligence/Machine Learning in its realization. We also instigate open research challenges for closed-loop automation of 5G and beyond mobile networks.