Samsung and KT Successfully Validate AI-RAN Optimisation on Live Networks

Samsung and telecom operator KT have successfully demonstrated the transformative potential of AI-optimised Radio Access Networks (AI-RAN) on live commercial networks. This landmark validation is the result of rigorous testing and represents the first time that this specific AI-driven architecture has proven its capabilities outside of controlled lab environments. By showcasing how AI can deliver stable service quality even in challenging real-world conditions, this trial highlights the shift towards more adaptive and user-centric network operations.

From Single-Configuration Networks to Dynamic AI Optimisation

The current surge in data consumption makes traditional approaches to network configuration increasingly unsustainable. Conventional cell towers apply a one-size-fits-all model, where a single configuration is assigned to every connected device. This rigid system often fails to respond effectively to varying connectivity scenarios—such as dense urban settings or high-mobility users—leading to frequent disruptions. As 5G networks continue to grow, these outdated methodologies become increasingly glaring bottlenecks.

The AI-RAN optimisation system, co-developed by Samsung and KT, addresses this limitation by using advanced predictive algorithms. Instead of optimising for an ‘average’ user, the AI system analyses real-time data on device conditions, movement patterns, and historical usage to apply customised configurations. This granular approach ensures that each device receives optimal service tailored to its specific needs, significantly reducing connectivity failures.

Successful Validation and Key Milestones

Samsung Research and KT’s Future Network Laboratory tested the AI-RAN architecture in Seongnam, Gyeonggi Province, focusing on users in regions with varying environmental challenges. The trial involved approximately 18,000 users, with a significant emphasis on those who had historically experienced frequent interruptions. By leveraging AI-driven predictive analysis, the system identified and resolved specific connectivity patterns that typically led to loss of service. The results were striking: not only did connection failures for targeted users decline, but the entire cell sector saw an improvement in performance, demonstrating the far-reaching impact of optimising edge cases.

This milestone serves as a crucial step toward the development of next-generation wireless networks. While 5G deployment is ongoing, the AI-RAN technology tested here sets the stage for the rollout of 6G, which will require handling massive data loads while maintaining reliability in increasingly complex mobility scenarios. Companies reliant on distributed IoT assets or mobile workforces can anticipate more stable service-level agreements (SLAs) as these advancements become mainstream.

Future Implications of AI-Driven Networks

The breakthrough achieved by Samsung and KT represents a seismic shift in network infrastructure. Wireless operations are increasingly becoming software-defined and predictive, with AI technologies enabling real-time adaptation to user environments. This evolution is key not only for individual users but also for critical industrial applications like IoT deployments and advanced mobile systems.

Experts across the industry are hailing this success as a step toward seamless global connectivity. As stated by JinGuk Jeong, Executive VP at Samsung Research, “This is a major milestone that shows how AI can improve user experience on real-world commercial networks.” Similarly, KT’s Executive VP, Jong-Sik Lee, emphasized that this achievement highlights the transformative potential of AI in delivering uninterrupted, stable service, paving the way for innovative 6G capabilities.

With AI-RAN optimisation, expectations from future connectivity providers are set to rise. Enterprises should evaluate the presence of AI-integrated solutions in a provider’s roadmap to ensure robust, adaptive services in high-demand or weak-signal settings.