MonB5G

Website:           Social media:            Contact:

Short abstract:

As 5G technology is maturing, and in view of new pervasive mobile services of different vertical industries, it will be a necessity to support massive numbers of coexisting network slices, with different performance requirements, functionality, and time spans. This puts significant strain on the management and orchestration system that traditional centralized designs, as in Cloud Computing and NFV, fail to cope with.

MonB5G answers to these challenges, towards providing zero-touch management and orchestration in the support of network slicing at massive scales for 5G LTE and beyond. It proposes a novel autonomic management and orchestration framework, heavily leveraging distribution of operations together with state-of-the-art data-driven AI-based mechanisms.

MonB5G is designed around a hierarchical approach that allows the flexible and efficient management of network tasks, while introducing a diverse set of centralization levels through an optimal adaptive assignment of monitoring, analysis, and decision-making tasks. MonB5G proposes a hierarchical, fault-tolerant, automated data driven network management system that incorporates security as well as energy efficiency as key features, in order to orchestrate a massive number of parallel network slices and significantly more diverse types of services in an adaptive and zero-touch way.

MonB5G reuses standards-based MANO and MEC frameworks, extending them with embedded cognitive capabilities. It further provides trust mechanisms tailored to the targeted multi-stakeholder environment, for secure and trustworthy cross-domain operations.

MonB5G has selected two use cases that will be trialed over the partners’ 5G testbeds, featuring automated, zero-touch slice management and orchestration across technical and administrative domains, enabling network operators to ensure end-to-end cross-domain SLAs, as well as AI-assisted policy driven security monitoring and enforcement.

Project Participants: