Exploring 5G Fronthaul Network Architectures

Wireless networks are evolving quickly as they experience ever-increasing traffic and as they connect to more devices. There is a pressing need for increased bandwidth and range as well as decreased latency and cost, as the industry prepares for the arrival and scaling of 5G networks. In many cases, networks are becoming less centralized and more complex, placing increased demands on the fronthaul network. In this paper, fronthaul refers to the connection from the cell site antenna to the central office where the baseband unit is housed, with the option for distributed units (DUs) with added intelligence processing closer to the antenna. To meet the need for increased capacity and low latency, while also managing costs, network architects and engineers face many options and trade-offs, particularly with respect to how much intelligence to place at the remote radio head (RRH) versus in a centralized location. There is no one-size-fits-all solution. Even within the same carrier network, different topologies may be deployed to meet the unique needs of different regions, and all of these are supported by Intel’s FlexRAN reference architecture. This paper presents and evaluates varying centralized versus distributed 5G fronthaul network topologies, the connectivity requirements for each, and explains where Intel® Silicon Photonics optical transceivers, an innovative combination of a silicon integrated circuit and a laser, offer an attractive solution to meet the high bandwidth and distance requirements associated with critical links in these topologies.