Dense user-centric RAN: Enhancing the capacity-energy efficiency trade-off in 6G

When developing radio technologies for the next generation of mobile networks (6G), energy efficiency must be a major criterion. In this talk, we highlight and discuss a promising approach and potential enablers to significantly increase throughput in 6G in an energy-efficient manner.  This challenge requires a radical reduction of the total energy consumption per transmitted bit, including the energy needed to run the (virtualised) hardware, to execute signal processing or AI algorithms, and to acquire the data for sensing and AI. We start with some theoretical results on scaling laws to gain valuable insights into the design of wireless networks that have the potential to operate close to the optimal trade-off between capacity and energy efficiency. On this basis, we argue for user-centric radio access networks (RANs) based on the idea of cell-free Massive MIMO (in the context of sub-6GHz and sub-THz communications).  In the case of sub-THz communication, which can also be used for fronthauling in the user-centric RAN, an energy-efficient design is mandatory. We therefore present and discuss a cross-domain approach that enables synergetic benefits to be created between hardware and baseband processing. The talk is rounded off with some considerations on intelligent reflecting surfaces and hybrid machine learning in the 6G air interface. The content of this talk draws on the research efforts of the 6G Research and Innovation Cluster (6G-RIC), which has launched an ambitious and comprehensive interdisciplinary research programme to address the technical challenges of 6G.