Keysight Technologies, Inc. (NYSE: KEYS) has joined forces with Samsung and NVIDIA to develop cutting-edge AI-driven Radio Access Network (RAN) solutions for 5G-Advanced and 6G technologies. This collaboration, showcased at Mobile World Congress 2025 (MWC 2025) in Hall 5, Booth #5F41, introduces an AI model integrated into Samsung’s vRAN software solution , promising transformative improvements in network performance and sustainability.
The Challenge: Limitations of Traditional RAN Systems
Traditional Radio Access Networks (RAN) face critical challenges in modern telecom environments:
- Limited Throughput : Especially at cell edges, where weak signals hinder performance.
- High Latency : Delays in data transmission affect user experiences.
- Inefficient Resource Utilization : Static rules-based algorithms struggle to adapt to dynamic conditions, leading to wasted energy and suboptimal network efficiency.
In low signal-to-noise ratio (SNR) scenarios, conventional channel estimation algorithms falter due to excessive noise, creating bottlenecks for operators.
The AI-Driven Solution: Precision and Efficiency
The collaboration focuses on AI-native RAN technologies to address these challenges. By training AI models using Keysight’s Channel Emulation Solutions , Samsung has developed advanced algorithms for uplink channel estimation in 5G-Advanced and 6G networks.
Key Innovations
- AI-Enhanced Channel Estimation :
AI models replace static algorithms, enabling precise signal processing even in low SNR environments. This improves cell edge throughput by 30% compared to traditional methods. - Energy Efficiency :
The AI model reduces power consumption while boosting performance, aligning with global sustainability goals. - End-to-End Testing :
The solution was validated using an NVIDIA GH200 Grace Hopper Superchip platform and Keysight’s Core Emulation Solutions , ensuring seamless integration into real-world networks.
Role of Each Partner in the Collaboration
- Keysight Technologies :
Provided Channel Emulation Solutions to simulate diverse radio frequency (RF) conditions and enable real-time signal processing. - Samsung :
Leveraged its vRAN software and radio points to implement AI models, focusing on improving uplink performance and network reliability. - NVIDIA :
Powered the AI pipeline with its AI Aerial platform , enabling flexible and scalable signal processing for next-gen networks. - AI-RAN Alliance :
The project is part of the alliance’s work to standardize AI-driven RAN technologies, ensuring interoperability and industry-wide adoption.
Lab Results: A 30% Leap in Performance
Simulated experiments demonstrated significant improvements:
- 30% higher cell edge throughput using AI-based channel estimation.
- Reduced power consumption while maintaining performance, critical for sustainable 5G/6G deployments.
These results were achieved through an end-to-end setup combining Samsung’s distributed unit (DU), NVIDIA’s AI Aerial platform, and Keysight’s emulation tools.
Quotes from Industry Leaders
- Charlie Zhang, Senior Vice President, Samsung Research :
“This collaboration with Keysight, NVIDIA, and the AI-RAN Alliance marks a milestone in creating AI-native networks . It addresses key challenges in uplink performance and sustainability, paving the way for smarter telecom systems”. - Soma Velayutham, Vice President, Telecoms, NVIDIA :
“AI unlocks unprecedented flexibility in signal processing. Our partnership with Keysight and Samsung accelerates the adoption of AI-driven RAN solutions , delivering efficiency gains for future networks”. - Giampaolo Tardioli, VP, 6G & Next-Gen Tech, Keysight :
“This project highlights the transformative potential of AI-native networks. By reducing power consumption and enhancing performance, we’re building scalable solutions for 5G-Advanced and 6G”.
Why This Matters for the Future of Telecom
The integration of AI into RAN marks a shift toward self-optimizing networks that adapt dynamically to user demands and environmental conditions. Benefits include:
- Improved User Experience : Faster data speeds at cell edges.
- Sustainability : Lower energy consumption aligns with global climate goals.
- Scalability : AI-driven infrastructure supports the complexity of 6G rollouts.
MWC 2025: Witness the Innovation Live
Visit Keysight’s Booth #5F41 in Hall 5 at MWC 2025 to see the AI-RAN solution in action. The demo will showcase how AI enhances network efficiency and prepares operators for 6G readiness.
Resources for Further Exploration
- Keysight Channel Emulation Solutions : Learn about the tools enabling AI model training for RAN 3.
- MWC 2025 Event Page : Stay updated on demos and sessions featuring AI-native networks 7.