Distributed Management of Network Slices in Beyond 5G
Overview
As 5G technology matures and new mobile services emerge across various vertical industries, supporting massive numbers of diverse network slices becomes essential. Traditional centralised management systems struggle to meet these demands, especially given Europe’s economic and societal shifts requiring digital transformation across sectors like automotive and e-Health. Network slicing is adopted to enable the coexistence of varied services with efficient resource use. However, challenges remain around scalability, security, automation, and energy efficiency. The move toward highly distributed, AI-driven management systems aims to ensure zero-touch orchestration, address trust between tenants and operators, and provide secure, privacy-aware monitoring in complex multi-domain environments.
To address these challenges, MonB5G developed a hierarchical, fault-tolerant, automated data-driven network management system designed to orchestrate a vast number of parallel network slices and diverse services in an adaptive, zero-touch manner. Key features included built-in security and energy efficiency. It aimed to enable zero-touch management and orchestration for network slicing at massive scales in 5G long-term evolution (LTE) and beyond, by leveraging a novel autonomic framework that distributed operations and employed advanced AI-driven mechanisms. Architecturally, MonB5G decentralised the traditionally centralised management system into multiple distributed sub-systems, spreading intelligence and decision-making across various components within different technological domains.
Objectives
Devise a distributed management plane to handle the deployment of a massive number of network slices.
Define network slice service-level KPIs that consider not only a single VNF, but all the network slice components, i.e., VNFs, PNFs and networking components.
Devise data-driven management system components (MS, AE, DE), based on SoA federated learning AI techniques.
Combine the Intent-based policy definition and the cognitive management entities, to target multi-domain zero-touch network configuration of sliced B5G networks.
Define decision algorithms tailored to the RAN. The envisioned decisions should allow the update of the RAN configuration, when necessary, to meet the heterogeneous performance requirements of multiple coexisting slices.
Elaborate and advanced AI-driven security mechanisms for automatic identification and mitigation of both in-slice and cross-slice attacks.
Provide ML-assisted techniques to optimise energy efficiency in all technological domains (i.e., Cloud, RAN, Core and MEC).
Develop dissemination, standardisation and exploitation of technologies in the MonB5G project. Special focus is given to push the solutions regarding the cognitive AE and DE to ETSI ZSM and ENI.
eBOS Role in the Project
eBOS played a key role in MonB5G by designing the security framework and developing AI-driven security algorithms to provide novel mechanisms for the mitigation of network slicing threats. It was also involved in integrating and testing all components, conducting proof-of-concept experiments, and led the dissemination efforts, including designing the project’s website. Through MobB5G, eBOS advanced its knowledge and understanding of security in 5G technologies, extending the company’s R&D activities. It applied the results in the company’s product portfolio to develop new, secure, innovative applications and solutions, expanding its customer base.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 871780.
