Nokia’s growing commitment to virtualized and open radio access network (RAN) technology is a point of pride for the Finnish juggernaut.

The RAN supplier anticipates the adoption rate of open RAN to follow Dell’Oro Group’s forecast of it reaching about 10% of the total RAN market by 2025, but it’s ambitions go well beyond that, Tommi Uitto, president of Nokia’s Mobile Networks business, told SDxCentral. 

A wave of development and innovation will have to be achieved before Nokia or any vendor begins transitioning in earnest to a fully open framework in the tightly controlled and largely proprietary wireless radio sector. However, Uitto is confident that Nokia and other players in the ecosystem have the talent and resources to bring this vision to fruition.

Nokia is increasingly contributing to and participating in groups like the O-RAN Alliance to make the technology a commercial reality someday, but it’s not yet on par with the current models, he said. This is also true of vRAN, which “is not yet as competitive a technology as bare-metal based radios,” but it will be too, eventually, he said.

Open RAN and vRAN today aren’t competitive on capex, opex, power consumption, or performance, according to Uitto. “But we know from overall technology development that the day will come when it is mature and it is competitive against bare metal,” he said. 

“It’s better to be inside and designing it, planning it, because that exposes us to all the experimentation. You learn a lot of stuff that you don’t by simply doing desk research. This way we can also make sure that our system-on-a-chip (SoC) and our platforms are O-RAN compliant, and that we also have the right solutions for vRAN,” Uitto said. 

A shift of this order is a complicated balance for any entrenched and well established business like Nokia to pursue. It remains to be seen how much a future RAN framework with more open interfaces will negatively impact existing RAN businesses, but Nokia is undeterred by that concern because it foresees more wins than losses.

When, or if, open RAN becomes the dominant industry standard, Uitto said he believes Nokia will gain more than it loses by opening up interfaces in its hardware and RAN software. “This is actually very much in our DNA. Nokia likes competition. We like open ecosystems,” he claimed, adding that all of Nokia’s new RAN products are open RAN compliant.

Indeed, Nokia started an initiative about 20 years ago to open the baseband architecture with a split between basebands and radio, according to Uitto. The vendor, at the time of just the second generation of cellular technology, began to wonder if it should sell its basebands to a competitor and focus on radios, he said, adding that it didn’t fully prescribe to the need for RAN vendors to develop and sell both types of equipment.

“The industry wasn’t ready for it and some of our competitors made an interesting strategic move by establishing CPRI (common public radio interface), and they did that to fragment the market so that there would be two different standards. And, of course, the rest is history,” Uitto said. 

CPRI, and now the enhanced eCPRI for 5G and open RAN standards, remains proprietary with some open elements, but the operation and management protocols are still completely closed, he explained. 

“Here we are again, and we think that there’s a lot that this industry can do better with more innovation, more competition, more speed,” Uitto said.

Open RAN is not just about opening interfaces in eCPRI, but also many other key layers of technology including the RAN intelligent controller (RIC), which introduces artificial intelligence to RAN operations. “Many of these open RAN baseband suppliers, or all of the new ones, they of course have to resort to commercial-off-the-shelf hardware or x86 computing, and that today is not capex efficient, nor opex efficient, so it is not cheaper,” he said. 

“Operators would also not accept spectrally less efficient technology when the most expensive asset they have is spectrum,” Uitto said.

Moreover, he added, vRAN is about more than merely comparing custom silicon in layer 1 processing against x86 processors with hardware acceleration. “You have to see the forest for the trees,” he said.

New processor technologies will evolve to the point where, even if there are capex deficiencies in layer 1, “it could be that the overall benefits of cloud computing and the decentralized network architecture with open RAN radio units (RU), distributed units (DU), and central units (CU), the overall benefits of automation, could actually outweigh the capex deficiency if you just look at the layer 1 processing piece,” Uitto explained.

The biggest challenge facing open RAN today lies in making layer 1 processing capex and opex efficient, he added. There are also engineering and operational challenges because “when we disaggregate the base station, we actually have to re-aggregate it somehow to ensure not just interoperability, but the right network performance, network operating concepts, and also lifecycle management from release upgrades to support policies,” Uitto said. 

“These are all the kinds of challenges that engineers can solve, and have been able to solve for the last decade, so we don’t have to invent new laws of physics here.”