Optimizing allocation and positioning in a disaggregated radio access network
Future wireless communication infrastructures, starting from 5G, will operate their radio access networks (RANs) based on virtualized functions distributed over a crosshaul, i.e., a transport solution integrating fronthaul and backhaul. Optimizing the resource allocation and positioning of the virtual network functions of a virtualized RAN (vRAN) is crucial to improve performance. In this paper, we propose a new optimization model to deal with VRAN functions allocation and positioning that seeks to maximize the level of centralization. Our model explores several representative functional splits, including the fully distributed remote unit (UK), while taking into account the limit imposed by the communication paths between the crosshaul and the core network. We compare our model with a state-of-the-art solution and show how our approach improves the centralization level in most of the scenarios, even considering the limit imposed by the core infrastructure. Our model also provides higher number of feasible solutions in most of the cases. Additionally, we investigate the positioning of the central unit (CU) and show that its placement with the core infrastructure is rarely the best choice.