Some operators believe the momentum behind Open RAN gives them their best opportunity ever to shake up their supply chains, introduce new ideas to the 5G RAN, and reduce their dependence on a shrinking group of large equipment vendors.
And several governments or regional powers are equally fascinated by Open RAN as a way to build or expand a homegrown industry for 5G, and therefore generate not just new revenues, but global influence in a key technology. (It remains to be seen how the Chinese intellectual property and innovation that is feeding into the Open RAN ecosystem will be handled, given the overtly or implicitly anti-Chinese nature of some of the initiatives.)
All this is certainly generating a welcome burst of innovation around 5G, network cloudification and disaggregated, interoperable RAN architectures. It is many years since so many innovative companies from all round the world were seen focusing on a platform that has become the stronghold of a small group of giants and their hand-picked partners. However, the big issue lies in the timing. As we have often argued, getting a brand new and disruptive platform to the maturity required to deploy mission-critical networks that must support vast amounts of traffic is not a quick or easy task.
Conventional networks are built on decades of R&D and practical implementation experience. The potential disruption, of course, comes from the fact that operators are, slowly but surely, moving away from conventional integrated RANs to cloud-based ones. The established vendors have to completely re-architect their networks and so are (almost) at a common starting point with new entrants (leaving aside all those years of accumulated radio, roll-out and management expertise).
If operators are prepared to wait until at least mid-decade, and probably longer, to deploy cloud-based macro RAN, the timing may be good for Open RAN architectures to rise to those challenges, whether it is traditional or new suppliers that make the grade (or, in the more idealistic views of the future, a multivendor platform provided by a large number of interoperating suppliers). Moving to a 5G Standalone network with a cloud-native core, capable of supporting many use cases and traffic types, is the next big migration challenge for many operators, and they tend to be prioritizing it over moving to vRAN.
However, some operators want to accelerate the pace, apparently to light a fire under Open RAN efforts, and they are supported by governments in the USA, Japan, India, European Union, UK – and other countries or regions hopeful of taking the lead in building a new 5G RAN industry. Last week, the five major European operators – Deutsche Telekom, Orange, Telecom Italia, Telefónica and Vodafone – that recently signed a memorandum of understanding to promote Open RAN and a European ecosystem, produced their first document. This outlined the technical requirements they have for a large-scale Open RAN, which are predictably demanding. But the surprise was that they set a collective target of starting large-scale roll-out in 2022 – by contrast with previous statements from Orange, Vodafone and others, indicating that 2025 would be the first year to introduce Open RAN to the 5G macro network.
This may be a bold statement to try to accelerate progress and inject urgency into current efforts, whether by vendors or by government and academic testing labs. It seems unlikely that operators will be able to deploy Open RAN in macro networks next year (especially as the five signatories already have significant new build-outs of conventional 5G RAN). They may be eyeing a more open, cloud-based architecture as they migrate to 5G Standalone, but realistically, localized and greenfield networks such as enterprise platforms or rural extensions will be the initial way to put a brand new architecture through its paces.
Too ambitious a timeline actually risks derailing the process. US vendors are already calling on their government to provide more concrete help, such as testing labs, if they are to be able to get their equipment market-ready in the timelines envisaged by the (often unrealistic) government and regulatory committees. The UK government also had a committee to study the diversification of the supply chain, and has at least followed through with a new interoperability testing initiative called SONIC (SmartRAN Network Interoperability Centre).
The other risk of pushing a timescale that is unrealistic for a wide range of suppliers to meet in a coordinated way, is that it will place the emerging Open RAN firmly in the hands of the few companies that might be able to produce a carrier-grade network that implements most, if not all, of the required capabilities.
Nokia and Samsung will no doubt be rushing to tick the boxes on the operators’ wishlist, in a way that could exclude smaller, slightly slower suppliers from their supposedly open platforms. And the response of the hyperscalers will be interesting too. They all have interests in driving open platforms that will help them in two main ways – accelerating 5G roll-out so there are more consumers and enterprises able to use their services on high quality mobile broadband links; and generating new revenues from hosting the vRANs and vCores on their clouds. Dish has already decided to run its vRAN in the AWS cloud; Microsoft and Google are working closely with telcos and enterprises to offer hosted services on Azure or Google cloud.
Facebook, the initiator of the Telecom Infra Project, recently applied for FCC permission to test its own self-optimizing network (SON) technology on a 5G small cell network in CBRS spectrum – a clear example of how the web giants are starting to develop key enablers of the new open networks themselves.
Another interesting TIP development is its newly launched Open WiFi project, which highlights how much the Open RAN community, and the cellular industry as a whole, could learn from the way that WiFi has already developed a broad ecosystem with huge scale. Despite that, however, there are still lock-ins and performance gaps in various layers of the WiFi sector, even after two decades of evolution. Open WiFi aims to address some of those while also kickstarting some long-needed convergence of WiFi and cellular platforms.
But the wider point is that it takes a long time for an open platform to mature and tick all the performance boxes, and it is unrealistic to think that Open RAN can move significantly more quickly than Carrier WiFi did, for instance. It will make very strong progress with the right support, including operator- and government-backed labs and trials. There is no reason why there should not be a strong, interoperable and capable 5G Standalone Open RAN platform ready for large-scale deployment by 2025 or 2026. But to address all the requirements outlined in the European operators’ document – energy efficiency, enhanced uplink, full containerization, a RAN sharing management scheme, to name just a few – will surely take more than a year? The document is the best description so far of what large operators really will require from their vRANs and open RANs, and will help focus industry R&D efforts, but being over-ambitious on timescales risks disappointment and even backlash down the road.