Japanese telecom giant KDDI has piloted infrastructure partner Samsung’s AI-powered radio access network (RAN) optimization platform on a live 5G standalone (SA) network, resulting in measurable improvements in network download throughput.
The trial used Samsung’s RAN Speed Optimization (RSO) platform on KDDI’s network using 100 megahertz of time division duplex (TDD) spectrum in the 3.7 GHz band across 407 cell sites around Tokyo. The test showed a 31% increase in 5G downlink speeds during “peak hours” and up to 52% increase in downlink speeds in “dense urban areas.”
Samsung explained that the RSO platform, which is part of the vendor’s AI-powered CognitiV Network Operations Suite, uses AI-based predictive models to automatically analyze mobile site data and recommend parameters optimized for that specific location. This helps optimize the performance and resource usage of cell sites, which are tenants associated with broader AI-RAN development.
ABI Research noted in a report that AI-RAN will generate more than $6.1 billion in investment revenue by 2032, but most of it is not expected to start until 2029. However, the analyst firm also highlighted the lack of AI-RAN Alliance member carriers, which is “an expression of skepticism on the part of the industry regarding the near-term value of AI-RAN.”
“Real growth in AI-RAN will only occur when performance benchmarks are validated in the field,” wrote ABI Research analyst Samuel Bowling. “Operators need evidence that AI-RAN can deliver technical and financial outcomes at scale, with a justifiable cost model. Without it, adoption will continue to lag vendor enthusiasm.”
AI-RAN operations and collaboration
KDDI’s results with Samsung are similar to those of other carriers.
Earlier this year, T-Mobile US noted that its AI-RAN work with Ericsson had incrementally improved network efficiency. The initiative was made possible by integrating AI-controlled RAN services designed to improve 5G network efficiency into commercial trials across the carrier’s 5G-Advanced network.
Specifically, the pair moved their AI-native scheduler and link adaptation into a “large-scale commercial trial” on live 5G-Advanced network traffic. The software leverages a neural network running on Ericsson’s “optimized” hardware that can predict changing radio conditions in real time, and tests showed a 10% increase in spectral efficiency and 15% increase in downlink speed compared to “traditional rule-based methods.”
The latest trial also continued long-standing efforts between KDDI and Samsung, with the vendor being appointed to provide 5G SA cores to KDDI and work on deploying an open RAN system.
Just last week, KDDI partnered Samsung with Nvidia and Keysight Technologies to begin developing a high-fidelity RAN digital twin aimed at supporting AI-driven optimization and autonomous network operations in the 6G era.
The project is designed to create a virtual replica of real-world RAN conditions, allowing carriers and vendors to train, test, and validate AI models without disrupting live commercial networks. KDDI said the platform supports large-scale simulation of changing wireless environments, traffic patterns, and network scenarios, providing AI systems with a safer and faster development environment than experimenting directly on production infrastructure.
