Virtual Resource Consolidation in the Edge for 5G Networks

The shift of the radio processing to the cloud, through cloud Radio Access Networks (C-RAN) technologies and of the cloud processing to the edge, through edge computing, form the environment in which 5G systems are being implemented, fostered and transformed from a future technology to a mainstream one. By its nature, global optimization of the edge resource deployment cannot by easily performed considering the number and the diversity of the players that will be involved in the edge computing arena. As a result, building and maintaining more and more edge-located resources for serving radio and application data will eventually lead to increased cost and energy consumption and resource underutilization. One way to overcome this predicament, is through virtual resource consolidation, where separate but efficiently interconnected edge resources appear as a single computing entity, serving radio and application data. In this context, we present the Virtual Elastic Datacenters (VEDC) in the edge notion for 5G networks that can alleviate these issues. We also describe an Integer Linear Programming (ILP) based mechanism for the placement of baseband and application processing loads in a VEDC-based environment, and perform respective experiments. We show that through VEDC resource consolidation better quality services can be provided, while improving resource efficiency.