FREE6G_RadEdge - Machine learning-based cell-free networks for 6G – Radio Edge

FREE6G-RadEdge (within the coordinated project FREE6G) lays emphasis on innovations at the RAN and Fronthaul domains, that will exploit the potential of cell-free networking in future 6G networks. FREE6G-RadEdge will develop novel cell-free networking mechanisms that will allow to leverage the Access Point (AP) deployment in a cost-effective manner, by exploiting the application of the distributed processing cell-free concept. FREE6G-RadEdge proposes to disaggregate the traditional cell-free Central Processing Unit (CPU) in Distributed Units (DU) and a Central Unit (CU) in line with the 3GPP NG-RAN architecture. APs could be interconnected with the cell-free DUs via fronthaul connections, either wirelessly or through a bus interface. As far as the former is concerned, APs will be linked through novel hybrid MIMO mechanisms with beam-steering and beam-sharing support. Many innovative algorithms for optimized AP clustering, local coordination, energy efficiency and topology adaptation will be proposed to optimally exploit the channel and traffic correlation patterns. FREE6G-RadEdge’s cell-free innovations will be implemented and integrated with existing vRAN elements, likewise, opening the door to provide inputs to the O-RAN project. At the transport domain, standard Optical Ethernet technologies will be leveraged for the Midhaul interfaces, while the network will be fully aligned with the Fixed-Mobile Convergence (FMC) paradigm, allowing the Midhaul optical fibres to be flexibly shared with Fixed services.