Espoo – TECHz Magazine – News Desk
Nokia has secured widespread global operator momentum for its artificial intelligence-powered Radio Access Network (AI-RAN) platform as tier-one telecommunications providers shift from early evaluations to live field trials. Moving rapidly from an industry vision toward commercial reality, the technology is being validated across North America, Europe, the Asia-Pacific, and the Middle East. Major global operators advancing active proofs of concept and trials include A1 Group, Chunghwa Telecom, du, e&, Mobily, stc, TPG Telecom, and Zain Saudi, building on earlier foundational validation milestones conducted with carriers like NTT DOCOMO, T-Mobile US, SoftBank Corp, and Indosat Ooredoo Hutchison.
The platform seamlessly combines Nokia’s AI-native anyRAN software architecture with the NVIDIA Aerial RAN Computer platform. This integration abstracts traditional radio functions into a fully programmable software layer, effectively transforming widely distributed cell towers into software-defined edge AI infrastructure capable of running millisecond-level distributed applications alongside standard connectivity workloads. Backed by an expanded collaboration focused on advanced AI-RAN developer tools, the system handles the unique, rising uplink demands and complex traffic patterns generated by generative, agentic, and physical AI systems.
Initial operational data confirms that Nokia’s AI-powered radio innovations have already achieved a greater than 20% improvement in spectral efficiency. A continuous, software-driven development roadmap is structured to scale these capabilities exponentially over the coming years, targeting a 50% efficiency gain by 2027 and exceeding a 100% gain by 2028. This projected milestone will effectively double network data capacity using existing spectrum assets, giving service providers a clear pathway to maximize network economics and deliver new edge AI services without relying on capital-intensive hardware refresh cycles. Crucially, the platform achieves this massive performance leap while remaining within the same power consumption envelopes as existing legacy systems.
To mitigate rigid vendor lock-in and lower capital risks for global operators, Nokia’s unified platform remains fully Open RAN (O-RAN) compliant and supports three flexible hardware deployment paths. Operators can choose to integrate GPU-powered plug-in expansion cards directly into their existing Nokia AirScale baseband units, deploy high-capacity standalone AI-RAN nodes for greenfield or high-density edge clusters, or implement Cloud RAN alternatives on commercial off-the-shelf (COTS) servers via independent platform partners. The broader ecosystem strategy also incorporates programmable merchant silicon from industry partners like Marvell.
By embedding advanced AI algorithms deep within Layer 1 and Layer 2 network scheduling architectures today, the platform establishes the foundational architectural baseline for AI-native 6G networks. This software-defined approach guarantees an evolutionary, software-only upgrade path to the upcoming generation of wireless networks, allowing future applications like integrated sensing and 3D digital twins to deploy natively over existing cell sites. Following the conclusion of current field validations, Nokia’s AI-RAN platform is scheduled to reach standard commercial availability through a software subscription model starting in 2027.
