Nokia AI-RAN Enters Multi-Country Field Trials, Spectrum Efficiency Improved by Over 20%

en.Wedoany.com Reported - On September 16, Nokia announced the expansion of its AI-RAN network testing scope. Operators including A1 Group, Chunghwa Telecom, du, e&, Mobily, stc, TPG Telecom, and Zain Saudi are advancing proof-of-concept, laboratory testing, and live network trials, uniformly adopting NVIDIA Aerial RAN Computer and Nokia AI-native anyRAN software. Existing tests cover markets in Europe, Asia-Pacific, and the Middle East, and related trials previously launched by Japan's NTT DOCOMO are also continuing.
Nokia disclosed that the AI-RAN platform has already achieved over 20% spectrum efficiency improvement through AI radio algorithms. According to its product roadmap, the platform plans to increase spectrum efficiency gains to approximately 50% by 2027 and to over 100% by 2028, primarily through software upgrades to increase uplink and downlink capacity on existing spectrum.
The platform consists of Nokia anyRAN software and the NVIDIA Aerial AI-RAN computing platform, and can simultaneously support 4G, 5G, and subsequent network evolution. For existing AirScale networks, Nokia offers a GPU-accelerated AirScale capacity expansion unit that adds AI computing capability while retaining existing radio equipment; the platform also supports Cloud RAN deployment and is compatible with radio units compliant with Open RAN standards.
Nokia released this commercial AI-RAN platform in July of this year and identified three hardware deployment paths, including expansion of existing AirScale networks, AI-RAN dedicated nodes, and cloudified RAN architecture. The product adopts a software subscription model, allowing operators to continuously add AI algorithms and radio performance features through subsequent software versions without simultaneously replacing all baseband hardware. The first pilot deployments are planned to begin by the end of 2026, with commercial versions planned to be available in 2027.
In addition to the newly added operators, Nokia has previously conducted AI-RAN evaluations with T-Mobile, SoftBank, and Indosat Ooredoo Hutchison, and has advanced GPU-accelerated AI-RAN feature testing with BT, Elisa, NTT DOCOMO, and Vodafone Group. In August of this year, Indosat, Ooredoo Group, Nokia, and NVIDIA also launched the Zankore AI infrastructure platform, combining AI computing power with communications network infrastructure deployment.
The testing focuses include AI-assisted radio resource scheduling, spectrum efficiency optimization, network capacity improvement, and hosting radio communications and edge AI workloads on the same computing platform. At present, each operator's project remains at the stages of proof-of-concept, laboratory testing, and field trials, and has not yet uniformly entered large-scale commercial deployment. Nokia plans to launch platform pilots by the end of 2026 and begin offering commercial products in 2027.
Related Products

GYTA Type Non-Self-Supporting Aerial/Duct Optical Cable
TONGDING INTERCONNECTION INFORMATION CO., LTD.
Multi-function Marine Monitoring Low-altitude Detection Radar Sea-air Boundary Ocean Dynamic Environment Measurement
Chengdu Dixin Technology Co., Ltd.
Industrial and Commercial Point-Type Gas Detector
Jinan Benan Technology Development Co., Ltd.

G.652.D Wavelength-extended Non-dispersion Shifted Single-mode Optical Fiber
HONGAN GROUP CO., LTD.


Comprehensive Mining Automation Control System
Beijing Tianma Intelligent Control Technology Co., Ltd.

Mine Explosion-proof and Intrinsically Safe Industrial Ethernet Ring Network (10 Gigabit/Gigabit)
Chongqing Mas Sci&Tech Co., Ltd.










