Project description
Tuning radio access network technology for 5G
In use since the beginning of cellular technology, radio access network (RAN) has evolved through the generations of mobile communications. As demand for mobile services increases and diversifies with 4G and 5G, networks need to meet a growing variety of quality of service (QoS) requirements. The EU-funded BANYAN project will develop key tools for data-driven 5G RAN. It will coordinate an academic-industrial cooperation to address several issues, including the modelling and forecasting of macroscopic high-dimensional mobile traffic patterns observed at RAN for individual services, at multiple scales in time and space and the geo-locating and characterising of in-building mobile traffic patterns observed at RAN. It will also develop algorithms to proactively optimise network slicing formation and life cycle management and proactively allocate virtualised RAN resources to meet users’ demands. BANYAN will also be responsible for training early-stage researchers, who are urgently needed by various sectors.
Objective
As mobile services consumed by people and machines become increasingly diversified and heterogeneous, 4G/5G
networks are asked to meet a growing variety of Quality of Service (QoS) requirements. Network slicing, enabled by Network
Function Virtualization (NFV), is a promising paradigm to increase the agility and elasticity of the mobile network via logical
slices that can be formed and composed dynamically, so as to adapt to the fluctuations in the demands for different mobile
services.
A key enabler for network slicing is accurate data-driven models and the prediction of the spatio-temporal dynamics of the
mobile service traffic, which allow discovering knowledge relevant to the orchestration of slices and anticipating the need for
their reconfiguration. The need for effective data-driven slice management is especially critical in proximity of indoor Radio
Access Network (RAN), which must accommodate most of the volume and variations in the demand associated to each
mobile service and whose performance is crucial to user QoS.
The BANYAN project is designed to address major open issues towards the realisation of data-driven 5G RAN, as follows:
- Modelling and forecasting macroscopic high-dimensional mobile traffic patterns observed at RAN for individual services, at
multiple scales in time and space;
- Geo-locating and characterising in-building mobile traffic patterns observed at RAN;
- Designing data-driven strategies for the allocation of 5G RAN resources;
- Designing data-driven policies for the orchestration of 5G RAN resources to suit service requirements and dynamics via network slices;
- Coordinating outdoor and indoor heterogeneous networks to meet user QoS requirements.
To address the research objectives above, BANYAN pursues a tight academic-industrial cooperation, which will allow
developing key tools for data-driven 5G RAN, as well as properly training early-stage researchers who are urgently needed
by industry, academia, etc.
Fields of science
Programme(s)
Funding Scheme
MSCA-ITN-EID - European Industrial DoctoratesCoordinator
CB23 3UY Cambridge
United Kingdom
The organization defined itself as SME (small and medium-sized enterprise) at the time the Grant Agreement was signed.