Zero-touch service management (ZSM) for 6G network services

Although 5G has not shown all its potential, 6G foundations have already been devised by the research community. Besides supporting 5G services composed of end-user as well as vertical industry applications, which already constitute a challenge by reporting to 4G services, 6G expects other services that introduce further requirements in terms of latency, reliability, and data rate. 5G has anticipated this evolution by relying on the concept of network slicing, which leverages the flexibility provided by network softwarization to build network instances (virtual networks) tailored to the application or network service. Building on network slicing, 6G will support diverse application and service requirements using the same physical infrastructure. Accordingly, the increased number of services available in 6G will lead to a situation where a massive number of coexisting network services, with different performance requirements, functionality, and timespans, run in parallel. This puts significant strain on the management and orchestration system that traditional centralized designs, as in Cloud Computing and Network Function Virtualization (NFV) fail to cope with. Indeed, a network service is deployed on different technological domains, i.e., Radio Access Network (RAN), Core Network (CN), Cloud, and Edge computing domains. Each technological domain uses its own tools to orchestrate and manage the resources, which complicate the overall management process; hence a centralized solution cannot be envisioned. Moreover, the high number of managed objects involved in the Life-Cycle- Management (LCM) of end-to-end network service calls for autonomic and self-optimized management and orchestration mechanisms to reduce human intervention to the minimum, hence decreasing the reaction time to sensitive service degradation and avoiding human errors. To overcome these challenges, it is important that the management and orchestration system of network services combine a hierarchical and fully distributed solution to cope with the heterogeneity of the managed objects of different technological domains, with state-of-the-art Machine Learning (ML) techniques and Artificial Intelligence (AI) algorithms to ensure more autonomy towards Zero-touch Service Management (ZSM).

This tutorial aims to give insight into the challenges that arise for the orchestration and management of emerging 6G network services when deployed over different technological domains and possible solutions that will mainly rely on AI and ML to achieve the ZSM vision. The tutorial is divided into two parts. The first part is more theoretical and aims to introduce some visions on 6G architecture as well as 6G services requirements and the challenges that arise when orchestrating 6G network services. Then, it reviews existing approaches for 6G network service management that are based on ETSI ZSM and ENI standards and the H2020 European project MonB5G. The second part of the tutorial will illustrate through concrete examples how to achieve the concept of ZSM per technological domain and an end-to-end vision. Among the provided example, several have been implemented using real testbeds based on OpenAirInterface (OAI) and Kubernetes.