If one theme has dominated Wireless Watch over recent years, it is the radical change in the dynamics of the RAN market. This has been driven by several significant factors, including the expanded performance requirements of 5G use cases; the emergence of new service providers, especially for enterprise users; geopolitics and the impact on Huawei.
The most important factor of all has been the start of the migration of the RAN to be a cloud-based network in which all the network functions from Layer 2 upwards, and some of the Layer 1 processes, are run in software on cloud infrastructure. This shift is only just beginning, since the RAN is far more challenging to cloudify than IT functions or even the packet core, and will always rely on very high end processors as well as specialist radio components and antennas.
However, while elements of the RAN can, of course, never be implemented in software or fully commoditized, the start of vRAN deployments is coinciding with powerful support – from operators, smaller vendors and governments – for an Open RAN platform that will broaden the supply chain and break the dominance of a few large OEMs. The move to 5G Standalone and to vRAN will often be the triggers for adoption of new Open RAN architectures, our operator surveys indicate, and operators also hope that Open RAN may ease the path to a fully cloud-based 5G network.
As more operators embark on Open RAN trials or deployments – particularly those in emerging economies or with significant enterprise and rural projects – the ecosystem will broaden and platforms start to mature (see item about MTN and others below). This has the potential to shake up the supply chain by lowering barriers for newer or smaller vendors to address the macro network market, not just the small cells where they have often found their feet initially.
Of course, there is still a scenario where the established vendors, with the likely addition of giants Samsung and NEC, pledge support for open interfaces and succeed in keeping their leadership of major network deals, playing on their years of expertise and the comfort factor of trusting an incumbent. But this will only happen if the tier one vendors can radically modify not just their architectures, but their pricing, for the age of cloud-based networks. Rakuten, which has initiated so many Open RAN changes in its greenfield network in Japan, has thrown down its most challenging gauntlet yet to the traditional suppliers, by insisting it will publish pricing of Open RAN elements to achieve transparency for the first time (see item below).
Cloud-based RAN does not just open opportunities for new RAN software and hardware vendors, but for the companies providing the cloud infrastructure to support cloud-native centralized unit (CU) and distributed unit (DU) functions. Suppliers of servers, processors and accelerators have the opportunity to address the rarefied RAN market for the first time, but the most important shift in dynamics will be the value chain relationship between the hyperscalers and the operators. Microsoft Azure, Google Cloud and Amazon AWS are increasingly targeting operators with offerings from edge cloud development platforms and infrastructure services, to full network-as-a-service (NaaS) propositions, which are starting with the core, but may extend to the enterprise RAN and then, even, the macro RAN in some cases.
But there is one factor that dampens enthusiasm for the dramatic effect open cloud models may have on operators’ future 5G RAN costs and agility. This is that, as operators move to 5G Standalone and the sliceable 5G core, most of the critical intelligence that will enable new 5G revenue streams will be in the core. The 5G core is a massive leap in functionality and intelligence from the 4G evolved packet core (EPC), and in most cases will be deployed cloud-native from day one. And it is multi-access rather than tied to 5G or even cellular access alone – convergence with WiFi and fiber will be important to some operators’ expansion models.
On one hand, if the core becomes the primary enabler of new revenues and business models, an open RAN architecture will become even more critical to reduce total cost of ownership in the network element that is becoming the workhorse rather than the brains of the operation.
On the other, it will make the RAN vendors less strategic to the operators than the core suppliers (and in the cloud-native world, it will be far easier to select RAN and core from different vendors, or to assemble a core from multiple suppliers’ xNFs). It was notable, last week, that amid the excitement about Vodafone’s choice of non-traditional suppliers for its UK Open RAN, the operator also awarded major 5G core contracts – not to a similar group of challengers, but to Ericsson. The big vendors may have to cede some share of the RAN market that has sustained them for so long – but that may be counted an acceptable loss as long as they hold onto control of the coming wave of 5G core deployments.