The architectures of radio access networks (RAN)s are continually evolving as a result of various technical and economic drivers. We regularly receive enquiries about RAN architecture from our readers and one topic that keeps coming up is the evolution from D-RAN to C-RAN and, inevitably, beyond.
A fairly simplistic answer is that the distributed or D-RAN approach refers to the antennas, RF front end and base station equipment being co-located on the same site. Introduced around the time of 3G, this approach refined the architectural definition of parts of the BTS and where digital interfaces were placed – the remote radio head (RRH), and the baseband unit (BBU) became entities that could be moved apart from each other, albeit they were linked by cable on the same site still. This made site deployment more flexible and could reduce signal loss if the RRH was placed closer to the antenna.
An equally simplified description of the initial stages of C-RAN – centralised or cloud RAN – is that network functions or base station functions are located offsite from the antenna and RF front end equipment. This makes it possible to use the same processing and resources for multiple base stations. Instead of three sectors, you might control, say, six or ten sectors from the same BBU – enhancing effectiveness and cutting costs.
The downsides? When you go from D-RAN to C-RAN you replace backhaul with fronthaul – and networks will move towards centralising the radio function using a CPRI (Common Public Radio Interface)-based fronthaul. CPRI is resource-intensive. That means you need a bigger pipe; it’s not easy for MNOs to mix and match vendors with CPRI. At the moment open CPRI does not exist; the company that sells you a BBU will also sell you an RRU. Indeed, our CTO Simon Fletcher spent eight years as part of the CPRI standards group contributing to CPRI and the evolution towards evolved CPRI (eCPRI) standard which aims to overcome some of the challenges of CPRI. Other initiatives are also underway such as the O-RAN alliance to help define open, interoperable interfaces and move the mobile ecosystem towards fully harnessing the technical and economic potential of virtualisation.
The cost-saving opportunities should extend to forms of inter-operator sharing, but a clear regulatory environment for sharing isn’t available yet. And a lot of the sharing benefits don’t apply to private cellular networks, though the appeal of private networks is, in any case, more about deploying a network that is under your control.
Thus there are advantages and disadvantages to C-RAN – and these are by no means the only ones – but you can certainly save money by switching to C-RAN, assuming you have a good, high-capacity fronthaul.
And C-RAN brings us closer to the, probably inevitable, virtualisation of the RAN, further lowering operator costs and permitting the addition of new capabilities to the network more quickly. In fact, several operators have achieved significant advances in virtualisation of most of the core network; it is already a reality and not an aim.
As for Real Wireless, we are uniquely qualified to monitor and assess this network evolution on behalf of partners, clients and governments having followed and influenced the evolution of RAN architectures for over ten years. That doesn’t mean we claim to be able to predict with total accuracy the future of the radio access network – but we have for some time already been helping to map out the options and trade-offs for clients such as our work for Small Cell Forum under their virtualisation release and our ground-breaking assessment of revenues, costs and business cases in 5G virtualised networks under the 5G NORMA and 5G-MoNArch projects. We are also applying our techno-economic insights on future networks to our participation in other projects and initiatives like UK5G and AutoAir, and through our regular involvement in future-focused workshops and events like 5G World Summit, RAN World, WWRF and 5G Huddle. Finally, I am very pleased to be joining the Cambridge Wireless Virtual Networks Special Interest Group (SIG) team as a SIG champion and am looking forward to exchanging experiences on the practicalities of making software defined networking (SDN) and virtualisation a reality in our upcoming ‘Reality Check for SDN’ event on 30 October.
And, as a multi-disciplinary consultancy that includes engineers, economists, project managers, business leaders, regulatory strategists, industry analysts and, of course, wireless technology experts, we are, we believe, better positioned than most to read the RAN roadmap as the route ahead reveals itself.