Samsung Electronics Co., Ltd. (Headquarters: South Korea, hereinafter referred to as Samsung) announced today that it has successfully completed a proof of concept for radio network speed optimization using Samsung's AI solution, "RSO (RAN Speed Optimizer)," on KDDI's commercial 5G SA (Stand Alone) network in Japan. This joint proof of concept, conducted over several months starting in the latter half of 2025, demonstrated that the AI solution is effective in actual network operations and can contribute to significant improvements in network performance. The proof of concept, carried out in urban, suburban, and rural areas in and around Tokyo, utilized 100MHz of TDD frequency in the 3.7GHz band across hundreds of cells. As a result of training a robust AI model under real-world conditions with diverse traffic and network environments, Samsung's RSO increased 5G communication speeds by an average of 31% across the entire test area during peak hours, with a maximum increase of 52% recorded in urban areas. In addition to this significant performance improvement, the experiment demonstrated the potential and effectiveness of AI solutions in enabling more precise and adaptive network optimization in commercial environments. Samsung's AI solution RSO achieves network performance improvements through unique parameter optimization for each cell. This represents a significant change and advancement beyond traditional cluster-level approaches where AI adjustments apply the same parameter settings to multiple cells. Samsung's RSO uses proprietary AI-based predictive models to automatically analyze site environment data and recommend optimized parameters for each cell. RSO is a diverse AI solution and is part of Samsung's CognitiV Network Operations Suite (NOS), which consists of AI agents and intelligent applications. The proof of concept also confirmed that by automatically adapting to changes in different network environments, telecommunications operators can reduce manual o