The approach integrates data generation, artificial intelligence/machine learning (AI/ML) training, and performance benchmarking into a single automated workflow, aiming to accelerate innovation and reduce deployment risk for operators and manufacturers.
As AI becomes more deeply integrated into the RAN, engineers must validate algorithms under diverse, real-world network conditions. Data collection, model training, and performance evaluation are typically performed in separate environments, making it difficult to compare, reproduce, and trust results before deployment, especially for physical layer functions like channel estimation, which directly impact performance, reliability, and user experience.
To address this challenge, Keysight’s AI-RAN Simulation Toolset orchestrates a comprehensive workflow that automates realistic scenario and dataset generation, AI/ML model training, and repeatable benchmarking. This enables engineers to consistently compare approaches and gain clear performance insights before field testing. At MWC, this workflow is deployed end-to-end using partner computing and radio platforms to validate AI-based RAN behavior in a realistic, controlled environment.
The demonstration integrates Keysight’s AI RAN Simulation Toolset with the NVIDIA Aerial Testbed, a comprehensive over-the-air AI-RAN testbed that operates on platforms such as the NVIDIA GH200 and NVIDIA DGX Spark, along with the NVIDIA Aerial Omniverse Digital Twin and commercial radio hardware like Analog Devices’ Titan O-RU platform. The AI/ML models were developed by Samsung, NVIDIA, and Keysight, and are trained and evaluated within a unified test flow to deliver actionable performance insights.
Soma Velayutham, vice president of AI & Telecoms at NVIDIA, stated that validating complex algorithms under real-world conditions is critical as AI is integrated into the radio access network, and that a unified flow simplifies the data chain required to train, validate, and deploy AI-native 5G and 6G networks.
Charlie Zhang, executive vice president of Samsung Research America, emphasized that collaboration revolutionizes network deployment by bridging the gap between theoretical models and actual implementation, providing a solid and commercially viable foundation for the future of 6G networks.
Balaji Raghothaman, Chief Technologist-6G at Keysight, underscored that AI in the RAN only delivers value when it can be validated with confidence, and that this automated workflow helps operators and vendors accelerate deployments, reduce risks, and make informed decisions by introducing AI-powered RAN capabilities.
