A joint backhaul and RAN perspective on the benefits of centralised RAN functions

Abstract

Cloud-RAN (C-RAN) is considered a prime 5G enabler with promising resource pooling gains, tighter coordination among cells, and cost saving in remote radio heads and corresponding deployment and operation. However, C-RAN brings stringent requirements on the backhaul last mile, or the fronthaul, in terms of capacity, latency, and synchronisation, to the extent that direct fibre is believed to be the only plausible fronthaul solution. Knowing that more often than not, fibre to the home is not available and is a cumbersome and costly technology to provide, what are the alternatives for deploying C-RAN? How much loss is incurred in a 5G network if C-RAN is not available? On the other hand, the distributed RAN (D-RAN) is less demanding on the backhaul but is believed to lack in performance in terms of resource usage and efficiency of RAN deployment. In this work we address the comparison between C-RAN and D-RAN from a joint RAN and backhaul perspective in a quantitative manner, using a case study approach. Our results show that C-RAN is indeed cost effective and advantageous from a joint perspective; moreover, intermediate functional splits between the C-RAN and the D-RAN are promising as an evolution path towards 5G, in the absence of fibre.