KT and Rohde & Schwarz to Showcase AI-Enhanced Radio Transmission Performance

Why It Matters

The 50 % throughput gain illustrates how AI can dramatically improve network efficiency and user experience, accelerating the rollout of 5G‑Advanced and future 6G services. Validating cross‑vendor AI models also paves the way for open, flexible standards.

Key Takeaways

• AI-driven CSI feedback lifts downlink throughput by 50%
• Two-sided AI model enables cooperative CSI compression
• Multi-vendor AI interoperability validated for 6G standards
• Enhanced video streaming experience demonstrated at MWC Barcelona
• Rohde‑Schwarz CMX500 provides realistic test environment

Pulse Analysis

The convergence of artificial intelligence and radio access networks is reshaping the roadmap for 5G‑Advanced and the emerging 6G era. Central to this shift is channel state information (CSI) feedback, a data‑intensive process that determines how efficiently a base station can adapt its beamforming and scheduling decisions. Traditional CSI handling incurs significant overhead, limiting spectral efficiency. By embedding machine‑learning models directly into the feedback loop, operators can compress and predict channel conditions with far fewer bits, unlocking higher throughput and lower latency. This AI‑driven approach is rapidly moving from research labs to commercial trials, promising a more intelligent, adaptable wireless fabric.

At MWC Barcelona 2026, KT, Qualcomm, and Rohde & Schwarz turned theory into practice with a live showcase built around Rohde & Schwarz’s CMX500 one‑box tester. The setup paired a Qualcomm AI‑enabled device prototype with the CMX500 acting as an AI‑powered base‑station emulator. Using a two‑sided AI model for CSI compression, the system delivered roughly a 50 % boost in downlink throughput over a conventional non‑AI baseline. The performance uplift was immediately visible in a high‑definition video‑streaming demo, where buffering vanished and picture quality remained stable even under congested conditions.

The successful demonstration carries weight beyond a single booth. It proves that AI models from different vendors can interoperate in real time, a prerequisite for the open‑architecture standards that 6G regulators are drafting. For network operators, the promise of AI‑enhanced CSI means lower capital expenditure on spectrum and infrastructure while delivering richer services to consumers. Test‑equipment manufacturers like Rohde & Schwarz also stand to gain, as their platforms become essential validation tools for complex AI algorithms. As the industry races toward commercial 6G, such collaborations will shape the competitive landscape and accelerate standard‑setting timelines.