- Nokia is establishing an R&D center in Dallas to accelerate the development of AI-RAN with partners
- The Finnish vendor says AI-RAN provides a “transformative impact” on the future of telecom infrastructure and services
- T-Mobile’s Ulf Ewaldsson joined Nokia execs on stage to testify to the role it plays in the “un-carrier’s” network
MOBILE WORLD CONGRESS, BARCELONA – Nokia is stepping up its AI-RAN efforts, using its usual pre-show press and analyst reception Sunday to review a litany of initiatives it’s driving in order to empower the RAN with AI.
T-Mobile, KDDI, SoftBank and Nvidia are some of the industry heavyweights it’s working with to advance AI and the RAN. Under its “anyRAN” approach, Nokia says its evolving cloud RAN solutions to include AI computing in shared infrastructure to maximize operators’ resource efficiency.
Nokia was one of the founding members of the AI-RAN Alliance, which launched one year ago with SoftBank and Nvidia, as well as Arm, DeepSig, Ericsson, Microsoft, Northeastern University, Samsung Electronics and T-Mobile. The organization is now 75 members strong and growing.
Nokia announced that it's establishing an AI-RAN Center at its U.S. offices in Dallas, where it will work with industry partners to develop and test AI-RAN solutions in real-world network conditions with a focus on creating new use cases, prototypes and to validate AI-RAN reference architecture.
More than network performance
Nokia President of Mobile Networks Tommi Uitto said one of the most straightforward and obvious use cases for AI is to improve network performance.
But there’s much more to it than that.
Of course, AI applications are capable of running in end user devices, including frameless eye glasses like the ones he was wearing. “But you can’t put all of the AI computing capability in this type of device,” he said. “You actually benefit from splitting the AI computing between the end user device and the base station of the network.”
Say an operator in the United States has 90,000 the base stations and they’re all configured to handle the highest amount of traffic. There are times when not all that capacity is needed. What if the paradigm shifts and the base station site becomes more intelligent with AI? And what if that extra capacity were sold to whomever wants it?
Those are the kinds of things the AI-RAN Alliance is talking about when they’re describing the three components of AI-RAN: “AI for RAN” (enhancing spectral and energy efficiency); “AI and RAN” (enabling new services such as edge AI inferencing); and “AI on RAN” (running AI workloads on the network).
Well-timed, Nokia
One special guest Nokia invited to the stage was T-Mobile President of Technology Ulf Ewaldsson, who spoke about T-Mobile's lengthy and ongoing work with Nokia – and certainly not sounding like someone who’s ready to cut Nokia loose despite reports to the contrary.
Ewaldsson talked about the work T-Mobile is doing in the enterprise side of the house for large-scale network slice rollouts at hospitals, retail, airports and other verticals.
“We need to be able to do this much, much faster, therefore … we’re moving over to a cloud native infrastructure and our partnership with Nokia is absolutely essential to do that,” he said, noting that Nokia is providing the orchestration engine and other pieces for the cloud initiative.
As noted, T-Mobile is a founding member of the AI-RAN Alliance and it has been working with Nvidia and OpenAI to incorporate AI more tightly into its customer-facing and network systems.
Ewaldsson was asked which piece of the aforementioned three AI-RAN components is most important.
That’s hard to answer because they intersect and there are so many dimensions to it, he said, concluding that all three are essential.
Stay up to date on all the news from Mobile World Congress with Fierce Network and FNTV.