Keysight and MediaTek Advance AI‑Driven Uplink Optimization and Model Life Cycle Management for Radio Access Networks
Working prototype shows how artificial intelligence-radio access network techniques can adapt to site conditions and improve uplink performance at scale
Keysight Technologies, Inc. (NYSE:KEYS) and MediaTek announced a working prototype that advances AI‑driven uplink optimization and model life cycle management for next‑generation radio access networks (RAN). The prototype will be demonstrated at Mobile World Congress 2026 and shows how AI‑enabled, RAN-assisted intelligence can improve uplink performance in real time while remaining effective over time through practical model retraining and over‑the‑air (OTA) updates, helping operators and ecosystem partners accelerate the transition to networks.
As networks evolve, maintaining consistent uplink performance becomes increasingly challenging. Coverage conditions vary widely across urban, rural, indoor, and hard-to-reach locations. Traditional transmitter diversity techniques are often static, which can limit performance as conditions change. At the same time, AI models trained in one environment may not perform the same across all deployment scenarios and updating those models in live networks can be operationally complex. Operators need an approach that is accurate, site‑aware, and easy to maintain.
Keysight and MediaTek address this need by combining RAN‑assisted AI decision‑making, site‑specific model retraining, and over-the-air (OTA) model updates in a controlled test environment that mimics real-world network conditions. The prototype leverages Keysight's Channel Studio RaySim and its Network and Channel Emulation solutions to create realistic, repeatable validation scenarios. This enables measurable gains in uplink throughput, spectral efficiency, and reliability while reducing the operational effort required to support diverse deployment environments. The work aligns with the objectives of the AI‑RAN Alliance, supporting scalable deployment and continuous optimization of AI in the RAN.
Mingxi Fan, Senior General Manager of Wireless Technology Group, MediaTek, said: "We're looking forward to highlighting how instrumental uplink performance enhancement will be in scaling future mobile networks so they can support AI-era applications more efficiently. Transmit diversity is a promising technique to be implemented at the device level so it can work across a wide variety of network environments, and our collaboration with Keysight demonstrates that through open infrastructure and device cooperation, our technology will facilitate practical, scalable AI-native RAN deployments."
Peng Cao, Vice President and General Manager, Wireless Test Group, Keysight, said: "Together with MediaTek, we have demonstrated a novel AI-enabling-AI technique which helps address the increasing demand on mobile networks' uplink performance, capacity, and consistent end user experiences driven by AI and AR applications. This involves training and testing the AI-enhanced transmitter model in the UE with high-fidelity real-world training datasets generated from Keysight's RF digital twin channel emulation solutions with 3D ray tracing capabilities."
About Keysight Technologies
At Keysight (NYSE:KEYS), we inspire and empower innovators to bring world-changing technologies to life. As an S&P 500 company, we're delivering market-leading design, emulation, and test solutions to help engineers develop and deploy faster, with less risk, throughout the entire product life cycle. We're a global innovation partner enabling customers in communications, industrial automation, aerospace and defense, automotive, semiconductor, and general electronics markets to accelerate innovation to connect and secure the world. Learn more at Keysight Newsroom and www.keysight.com.
View source version on businesswire.com: https://www.businesswire.com/news/home/20260227733318/en/
Media Contacts
Andrea Mueller
Americas
[email protected]
Fusako Dohi
Asia
+81 42 660-2162
[email protected]
Jenny Gallacher
Europe
[email protected]