Keysight’s holistic network digital twin solution streamlines next-generation AI-native wireless communication systems, supporting the entire workflow from research and development (R&D) to planning, deployment, optimization, and standardization of AI / ML capabilities in 5G-Advanced and 6G systems. Future wireless networks will integrate AI / ML algorithms and neural networks at the physical layer, replacing traditional mathematical models for radio signal processing. Deploying these AI models in production systems will require powerful and precise end-to-end simulation platforms capable of testing, validating, and fine-tuning AI models using both real-world data and synthetic data representing thousands of site-specific scenarios.
In the future, AI models will be simulated and optimized in the digital twin before being deployed in the physical network, transforming digital twins from post-deployment optimization tools into essential design and validation environments. This is where advanced platforms like Keysight network digital twin will drive breakthrough innovations in AI-for-RAN, paving the way for AI-native 6G networks.
The Keysight network digital twin integrates extensive emulation solutions, including RF propagation modeling, base station, mobile, O-RAN, and core network emulation. It will accelerate AI-RAN adoption and commercialization, benefiting network equipment and device makers, wireless chip vendors, mobile network operators (MNOs), academic and research institutions, and others.
What Are the Benefits of a Network Digital Twin?
Network digital twins are becoming integral to the wireless industry, transforming network planning, optimization, and management. In telecommunications, they serve as a powerful tool for simulating and analyzing network behavior, allowing operators to anticipate traffic patterns, optimize resource allocation, and enhance overall service quality. In modern networks, digital twins facilitate virtual network slicing, enabling the design, deployment, and management of tailored service levels for diverse applications. For wireless infrastructure providers, digital twins streamline the design and testing of new network topologies, ensuring efficient deployment while minimizing the risks associated with rolling out new technologies.
By offering a dynamic, real-time view of the network, digital twins empower predictive analytics and fault detection, fostering more resilient, agile, and cost-effective wireless services. Researchers and academics are also harnessing the power of digital twins, making them a vital asset in driving innovation and efficiency across the wireless ecosystem.
Figure 1. Keysight RaySim RF propagation digital twin in PROPSIM emulation workflow
At Mobile World Congress 2025, we unveiled RaySim, our cutting-edge raytracing solution for highly detailed and accurate channel generation, marking a key milestone in the evolution of our next-generation digital twin platform. RaySim is a state-of-the-art RF raytracing tool designed to rapidly and accurately create site-specific RF propagation scenarios and channel models. It supports link-level, system-level, and network-level RF scenario simulations for both Keysight and customer digital twins, enabling real-time emulation for hardware-in-the-loop (HiL) testing through our PROPSIM channel emulator, as illustrated in Figure 1.
RaySim facilitates the seamless virtualization of drive test routes with high RF fidelity, enabling realistic and continuous integration testing of new software releases and network configurations. The solution supports MIMO and Massive MIMO radios, multiple UEs, and antenna beam pattern embedding, with wireless system parameters easily adjustable to evaluate various scenarios quickly in the lab.
An ideal tool for generating site-specific datasets, RaySim will accelerate AI / ML model training and benchmarking for AI-RAN and AI-Device use cases. RaySim leverages the NVIDIA Aerial Omniverse Digital Twin platform to enable modeling of wireless deployment scenarios in complex 3D environments with high fidelity. Using the computational power of NVIDIA accelerated infrastructure, RaySim accelerates complex RF raytracing, delivering highly accurate, site-specific channel models tailored for 6G R&D.
Figure 2. Helsinki city center urban microcellular (UMi) field measurement scenario with two cells implemented in Keysight RaySim
At NVIDIA GTC 2025, Keysight is showcasing a network digital twin proof-of-concept (PoC) powered by a high-fidelity RaySim-generated channel model, depicted in Figure 3, which is integrated with a 5G protocol stack and capable of running real application traffic. This event also includes an industry panel session titled, “Driving 6G Development With Advanced Simulation Tools.”
Figure 3. Video Streaming over Keysight Network Digital Twin
The fidelity of the digital twin is a critical factor in determining its effectiveness as a simulation and testing tool. To validate the accuracy of the digital twin, we conducted several RF field drive tests comparing predicted data from the digital twin with real-world data from live networks, as shown in Figure 4. The results demonstrated a high level of correlation, reinforcing the model’s precision and reliability.
Figure 4. RSRP comparison between field measurements and Keysight RaySim predictions
Keysight announced a collaboration with NVIDIA in March 2024 to leverage the NVIDIA Aerial AI platform to accelerate radio access network (RAN) design workflows. Since then, the partnership has evolved and strengthened, driven by increasing industry momentum for AI-native wireless solutions. This collaboration combines Keysight’s expertise in physical modeling, simulation, and emulation with NVIDIA’s strengths in accelerated computing, setting the foundation for next-generation wireless network innovation.