We have strong expertise in resource management & orchestration of multi-provider/-domain intent-based cloud/edge networks. This includes workload + traffic management to provide holistic SLA compliant optimisations across multiple domains. Objectives include (but not limited to):
- energy utilization,
- awareness of energy source types (green, brown,...) and maximization of green energy usage,
- resource scarcity/disparity
Our work also focus on scalable service discovery and resolution and efficient forwarding/routing solutions. Examples of our most recently published work include:
- https://ieeexplore.ieee.org/document/9910022/
- https://ieeexplore.ieee.org/abstract/document/9964993
We are interested in contributing to B-01-01 with 6G architectural solutions that for:
- Intent-based frictionless orchestration of cloud-to-edge inter-cluster networks, with intent space extending to AI and energy utilisation
- Energy efficient enablers: integration of scalable and decentralised optimisation algorithms and orchestration/control services that maximises use of green and renewable energy for cloud-edge workload and traffic allocations. Optimisation includes applied federated machine learning for demand forecast and ecosystem state prediction.
- Digital twining: 6G-ecosystem service-based monitoring-logging for efficient observability of end-to-end operations
The Network Convergence Laboratory (NCL) is one of three research laboratories, part of the larger COMNET group headed by Prof. Martin Reed, University of Essex. The NCL team has a rich research portfolio focusing on future Internet architectures, Fog computing, programmable networks, cybersecurity, multimedia communications systems, communications and networks for 6G and beyond, and the application of machine learning in these areas. NCL also maintains a well-equipped research network testbed that consists of programmable electrical and optical networking technologies and heterogeneous wireless access networking technologies across the University campus and surrounding areas. Over the programmable network, NCL host a hierarchical Fog computing substrate with 200+ heterogenous compute-storage nodes, distributed between a resource capable central cloud and multiple resource constrained edge computing islands. With that, NCL also offers an access network of 40+ nodes, emulating end-users’ usage patterns. NCL is also connected to GENI in USA over a software defined connection. These facilities have been used in various EU and national EPSRC projects, such as EPSRC NIRVANA and TANGO and EU H2020 POINT, iCIRRUS, RECENT and SerIoT.