The future 5G Internet has demanding new efforts from telecommunication operators to meet its demands. Cloud radio access networks (CRAN) based architectures have already been deployed by operators to cope with the large user coverage demanded by those networks, maintaining an energy-efficiently operation by centralizing the mobile data processing. However, CRAN demands high transfer rates on the fronthaul interconnecting the radio elements and the cloud.
Moreover, it introduces strict time delay constraints, which has led to the proposal of hybrid architectures that implement centralized and distributed closer-to-the-user processing using Fog Computing. In this context, the use of optical networks is still crucial to this communication due to its high transfer rates.
So, we propose an architecture called Cloud-Fog RAN (CF-RAN), that comprises the processing of mobile data in the cloud or in local processing nodes under a TWDM-Passive Optical Network as its transport interface in order to cope with the delay and transfer demands.
Experiments with a preliminary mechanism to schedule transmissions, which activate on demand the processing nodes and distribute the processing functions, brings energy saving up to 95.5% in comparison with the traditional distributed RAN and greater users coverage than CRAN.
Prof. Daniel Batista (firstname.lastname@example.org)
Prof. Edmundo Roberto Mauro Madeira
Prof. Carlos Kamienski
Prof. Gustavo Bittencourt Figueiredo
Prof. Massimo Tornatore
Prof. Biswanath Mukherjee
Rodrigo Tinini, PhD Student, IME-USP