Rohde & Schwarz Advances AI-RAN Testing Using Digital Twins with NVIDIA

Why It Matters

The testbed shortens the path from AI‑RAN research to commercial deployment, giving telecom vendors faster, more accurate validation while positioning Rohde & Schwarz and NVIDIA as leaders in next‑gen wireless testing.

Key Takeaways

• R&S testbed uses NVIDIA Sionna with DGX Spark.
• Enables hardware‑in‑the‑loop AI‑RAN validation at system level.
• AI link adaptation learns site‑specific and UE behavior.
• Supports 5G‑Advanced and 6G research without field trials.
• Demonstrated live at MWC Barcelona 2026 booth 5A80.

Pulse Analysis

Artificial intelligence is reshaping radio access networks, but the industry still lacks a reliable bridge between simulation and real‑world performance. Digital‑twin technology, powered by high‑fidelity ray‑tracing, offers that bridge, allowing engineers to model complex propagation environments without costly field measurements. Rohde & Schwarz’s partnership with NVIDIA leverages the Sionna Research Kit, a differentiable simulator that can generate synthetic channel data at scale, addressing data sparsity and privacy concerns that have hampered AI‑RAN training.

The new testbed fuses Sionna’s software‑defined 5G NR stack with a single NVIDIA DGX Spark, delivering real‑time AI inference under strict latency constraints. Integrated with R&S’s SMW200A vector signal generator and FSW spectrum analyzer, the hardware‑in‑the‑loop setup emulates site‑specific channels and feeds them to an AI‑driven link‑adaptation algorithm. This algorithm continuously optimizes downlink modulation and coding, learning both the environment’s propagation characteristics and individual user‑equipment behavior, thereby demonstrating true end‑to‑end system verification beyond link‑level proofs.

For telecom operators and equipment manufacturers, the ability to validate AI‑enhanced base‑station features in a controlled lab reduces reliance on expensive over‑the‑air trials and accelerates time‑to‑market for 5G‑Advanced and future 6G solutions. The collaboration also strengthens Rohde & Schwarz’s portfolio of test‑and‑measurement tools, while showcasing NVIDIA’s leadership in AI‑centric simulation. As networks become increasingly software‑defined and AI‑centric, such integrated testbeds will become essential for ensuring performance, security, and regulatory compliance across the next generation of wireless infrastructure.