Samsung Electronics and KDDI completed a multi-month trial of an AI-powered radio access network optimization system on KDDI’s commercial 5G Standalone network in Japan, demonstrating average downlink throughput gains of 31% during peak traffic periods. The trial began in late 2025 and covered hundreds of cells across dense urban, suburban, and rural environments in and around Tokyo.
The companies tested Samsung’s RAN Speed Optimizer (RSO) using 100 MHz of 3.7 GHz TDD spectrum. Samsung reported that RSO increased average 5G downlink throughput by 31% across the trial area during peak hours, with maximum gains of 52% in dense urban locations. The system uses an AI-based prediction model to analyze site conditions and recommend optimized radio parameters for each individual cell rather than applying common settings across clusters of cells.
RSO forms part of the Samsung CognitiV Network Operations Suite, which includes AI-powered automation applications, AI agents, and network operations tools. Samsung and KDDI plan to continue evaluating AI-based optimization technologies for broader commercial network applications, building on their existing work in fully virtualized mobile network deployments.
• Trial conducted on KDDI’s live commercial 5G Standalone network in Japan
• Deployment covered hundreds of cells across urban, suburban, and rural environments around Tokyo
• Network configuration used 100 MHz of 3.7 GHz TDD spectrum
• Samsung reported a 31% average increase in 5G downlink throughput during peak hours
• Maximum throughput improvement reached 52% in dense urban areas
• RSO optimizes radio parameters independently for each cell rather than applying common settings across cell clusters
• The AI prediction model analyzes site environment data and recommends cell-specific parameter configurations
• RSO operates as part of the Samsung CognitiV Network Operations Suite
• Samsung and KDDI plan further evaluations of AI-based optimization technologies for commercial networks
“Combining KDDI’s accumulated expertise in network innovation and Samsung’s technical leadership, this field trial proves that individual tuning for cells — a long-standing industry challenge — has now become a reality through the integration of AI,” said Kazuhiro Furuhata, Chief Network Officer at KDDI. “Moving forward, we remain committed to pushing the boundaries of AI-based technologies to continuously elevate network experience for our customers.”
🌐 Analysis: The trial highlights a shift in mobile network automation from centralized optimization policies toward increasingly granular, cell-specific control loops driven by AI models. Applying optimization at individual-cell scale could become more important as operators manage heterogeneous 5G Advanced networks, Open RAN deployments, massive MIMO configurations, network slicing, and eventually AI-native 6G architectures.
Samsung and KDDI already collaborate on virtualized RAN and Open RAN deployments, giving the companies a commercial network environment for evaluating AI-driven operations technologies. The broader telecom equipment market is moving in the same direction, with Nokia, Ericsson, and other network suppliers expanding AI-based RAN automation, autonomous network operations, and AI-RAN initiatives as operators seek to improve spectrum utilization and reduce the operational complexity of increasingly software-defined mobile networks.