Keysight and Samsung advance AI-for-RAN based on the NVIDIA AI Aerial platform

Keysight Technologies has partnered with Samsung and NVIDIA to develop an Artificial Intelligence (AI) model aimed at enhancing Radio Access Network (RAN) performance. This collaboration is set to improve throughput, increase system capacity, and reduce power consumption in Samsung’s 5G-Advanced and 6G technologies. The AI model will be demonstrated at Mobile World Congress 2025.

The use of AI modeling addresses challenges faced by operators in traditional RAN setups, particularly related to uplink performance and resource utilization. AI enables accurate channel estimation and optimized resource management, significantly improving network efficiency and user experiences. Samsung’s experimental results indicate a potential 30% increase in cell edge throughput when using AI for channel estimation.

Charlie Zhang, Senior Vice President at Samsung Research, stated that the collaboration represents a significant milestone toward AI-native telecommunication networks. Soma Velayutham, Vice President at NVIDIA, noted the increased flexibility and performance that AI brings to the radio signal processing pipeline. Giampaolo Tardioli, Vice President at Keysight, highlighted the impact of this partnership on energy-efficient networks.

Looking forward, this project signifies progress toward the implementation of AI-enhanced RAN technologies, providing a framework for future innovations and developments in wireless networks.