Skip to main content

End-to-End Cognitive Network Slicing and Slice Management Framework in Virtualised Multi-Domain, Multi-Tenant 5G Networks

Periodic Reporting for period 2 - SLICENET (End-to-End Cognitive Network Slicing and Slice Management Framework in Virtualised Multi-Domain, Multi-Tenant 5G Networks)

Reporting period: 2018-07-01 to 2020-06-30

SliceNet has implemented an innovative framework for management and control of Network Slices (NSs), leveraging Software-Defined Networking (SDN) and Network Function Virtualization (NFV) technologies, with cognitive techniques and artificial intelligence for 5G networks. SliceNet overall achievements were:
- Achieve an innovative, cognitive, integrated ‘one-stop shop’ 5G slice management framework for vertical businesses and co-designed by vertical sectors:
Integrated SliceNet network slice control, management and orchestration framework for multiple business roles
One-Stop API to allow multiple business roles to access the SliceNet framework to subscribe, instantiate and optimise network slice based services
P&P (Plug & Play) Control to allow exposure of network slicing monitoring and control facilities to verticals to achieve service customisation for verticals
- Enable extensible, end-to-end slice FCAPS (Fault, Configuration, Accounting, Performance and Security) management across multiple planes and operator domains:
End-to-end FCAPS management of network slices in the integrated SliceNet framework for both single and multiple domains
Versatile P&P management framework that provides automated deployment and configuration for per-slice/per-service functions serving the verticals, and also the functions in SliceNet control plane and cognition plane
- Establish cognitive, agile QoE (Quality of Experience) management of slices for service assurance of vertical businesses:
Cognitive network management framework across Verticals, Digital Service Providers, and Network Service Providers (5G Operators)
Cognition-based artificial intelligence algorithms for QoS and QoE optimization
- Enable slicing-friendly infrastructure and coordinated, true provisioning and control of user-definable slices:
ETSI standard-compliant MEC (Mobile/Multi-access Edge Computing) platform and services
Programmable data plane for network slicing over the data path to assure QoS
Slicing-friendly end-to-end infrastructure that allows the SliceNet control plane to optimise/customise network slices at runtime
- Empower orchestration for cross-plane coordination of management, control, service and data planes to achieve system-level slicing control and slice operation:
Multi-level network slice orchestration across resource, slice and service levels
Automatic deployment of end-to-end network slices across multiple network service providers’ domains
- Demonstrate the efficiency and support of the SliceNet framework via the use of the vertical business use cases for validating the technical developments and the capabilities and flexibility of the integrated SliceNet framework
Showcasing use cases representative of the 5G KPI categories defined by 3GPP/ITU
Validate and showcase the technical, societal, and business impacts of the SliceNet system throughout the use of the developed vertical prototypes
- Create a 5G Slicing themed innovation ecosystem
Use cases testbeds for the three vertical sectors: e-Health, Smart Grid, and Smart City
End-to-end multi-domain multi-tenants network slicing framework that fosters collaboration among various business roles
Vertical-in-the-loop R&D based on requirements from the representative verticals in the project
"The main outcomes of the work demonstrate the project achievements beyond the state of the art, and the project potential to create impact both for ICT and for the verticals:
- SliceNet commercialisation opportunities:
The eHealth use case features practical tools for paramedics and other healthcare workers including wearable camera headset for ambulance crew or pandemic fighting scenarios, machine learning based stroke diagnosis and video streaming optimisation software, among others.
The Smart Grid use case integrates with existing Smart Grid Intelligent Electronic Devices (IEDs) for self-healing capabilities against grid faults/failures.
The Smart City Lighting use case has been deployed in real world to save energy for a university campus in Bucharest Romania.
- SliceNet impact on standardization and open source:
ML5G-I-247 – Machine learning based end-to-end network slice management and orchestration
ML5G-I-242-R1 – Vertical-assisted Network Slicing Based on a Cognitive Framework,
ETSI ENI POC (ENI(19)000055) “Predictive Fault management of E2E Multi-domain Network Slices"""".
OpenAirInterface (OAI),
Mosaic5G,
OpenvSwitch (OVS),
SkyDive,
OpenWhisk.
- Contributions to the 5G PPP program:
AI-empowered cognitive network management & orchestration, under 5G PPP TB Golden Nuggets List #7. 5G Network Management
End-to-end multi-domain multi-tenants network slicing, under 5G PPP TB Golden Nuggets List #8. 5G Multi-domain multi-tenants Plug & Play Control Plane and Slicing control
Plug & Play control & One-stop API for verticals’ slice customization, under 5G PPP TB Golden Nuggets List #8. 5G Multi-domain multi-tenants Plug & Play Control Plane and Slicing control"
SliceNet architecture has been successfully integrated and the system’s closed-loop successfully validated.
SliceNet framework allows building and managing network slice based services to meet the requirements of business use cases. Three use cases have been successfully demonstrated over the integrated SliceNet framework, well aligned to the 3GPP/ITU categories: eMBB for eHealth, URLLC for Smart Grid, and mMTC for Smart City, and thus very representative for a wide range of different use cases.
The SliceNet project concluded its 37 months of lifetime, having successfully achieved the objectives and generated innovative outcomes with great potential for creating impact.
The overall achievements of the project, which have culminated during the third year period, are the SliceNet framework fully integrated and applied to the three use cases (eHealth, Smart Grid, and Smart City), with proof-of-concept and key performance evaluation conducted via representative use case scenarios based on the requirements from the vertical industries.
Notable achievements have been made in dissemination and exploitation in the project related to the above results. Key achievements comprise the following highlights:
- 60+ papers in international journals, magazines or conferences including several top-tier journal publications.
- Numerous dissemination events and materials organized or produced including live recorded webinars for all the technical work, contribution to white papers through liaison with 5G PPP program, among others.
- 7 formal contributions to international standardization organizations including 5 official Work Items in ITU-T ML5G and ETSI GANA, a ETSI ENI POC and an IETF draft contribution.
- Contributions to different open source communities
- Close collaboration with other 5G projects, especially joint development and trial with Phase 2 project MATILDA which has been deployed and demonstrated in real-world operational environment.
It is noted that the project has significantly exceeded the dissemination and exploitation KPIs defined in the Description of Work.