T-Mobile and Ericsson have reported performance gains from trials of an AI-native radio access network feature running on T-Mobile’s 5G Advanced network.
The companies said the feature has moved into large-scale commercial trials using live 5G Advanced network traffic. Compared with legacy rule-based methods, Ericsson’s AI-native Scheduler with Link Adaptation delivered nearly a 10% improvement in spectrum efficiency and up to a 15% increase in downlink throughput, consistent with earlier tests conducted in smaller areas.
Trials began in early Q2 2025 and expanded to multiple U.S. markets, including Los Angeles, New York, New Jersey, and Salt Lake City. T-Mobile is targeting commercial deployment in Q3 2026.
Scheduler and link adaptation
The tested software embeds AI into the RAN scheduler and link adaptation processes, using a neural network running on Ericsson hardware to predict evolving radio conditions. It is designed to make real-time scheduling and link adaptation decisions as radio conditions change, including in high-traffic scenarios or where radio frequency conditions are weak.
In practical terms, the scheduler controls how radio resources are allocated among users, while link adaptation adjusts transmission settings based on signal quality and interference to maintain efficient communication.
Ericsson says the feature aims to deliver more stable performance for services such as streaming and video calls during periods of heavy network use. The AI runs inside the RAN software as well as in external network management tools to provide faster, context-aware decisions.
Johan Hultell, head of Product Line RAN Software for Business Area Networks at Ericsson, emphasized that AI is central to the company’s strategy for programmable, adaptive networks.
NVIDIA infrastructure work
This activity is part of T-Mobile’s broader AI-RAN work with Ericsson, Nokia, and NVIDIA. In September 2024, T-Mobile opened the AI-RAN Innovation Centre at its Bellevue, Washington headquarters to collaborate with these partners.
Ericsson has also demonstrated Cloud RAN software on NVIDIA AI infrastructure with T-Mobile at the Bellevue centre. That Cloud RAN software is built to run on commercial off‑the‑shelf hardware and NVIDIA AI systems, and it can also run on purpose-built platforms that use Ericsson silicon.
When deployed on general-purpose hardware, the software can leverage available accelerator hardware and software for compute-intensive RAN functions.
Ericsson said the demonstration with T-Mobile used NVIDIA Aerial CUDA. That work is separate from the scheduler trial and focuses on Cloud RAN software portability and performance on different hardware platforms.
T-Mobile is using the Bellevue centre to validate AI-RAN and Cloud RAN deployments alongside Ericsson, Nokia, and NVIDIA. Those four companies are also founding members of the AI-RAN Alliance, launched at Mobile World Congress Barcelona in February 2024 to explore how AI can be integrated across cellular networks, including RAN and mobile edge systems.
Ericsson and T-Mobile said they will continue collaborating on additional ways to apply AI to improve network performance and operational efficiency.
Want to discover how IoT is transforming telecoms and connectivity? Attend the IoT Tech Expo in Amsterdam, California, or London to learn how 5G, edge computing, and IoT innovations are shaping network and service evolution. The event is part of TechEx and co-located with other technology conferences.
Telecoms Tech News is produced by TechForge Media. Explore other upcoming enterprise technology events and webinars through TechForge Media’s events listings.