KT and Samsung Cut High-Speed Rail Connection Drops by 29 Percent Using AI

KT and Samsung Electronics have developed a new artificial intelligence technology that reduces communication dropouts on high-speed railways by approximately 29 percent. The two companies announced on the 5th that they successfully tested the wireless network optimization technology, which adjusts base station handover timing using AI, along a section of the Gyeongbu Line.

Operating at speeds exceeding 250 kilometers per hour, high-speed trains experience rapid shifts in base station signals, requiring smartphones to switch connections frequently. When these handovers fail to occur at the precise moment, data transmission is delayed and call quality deteriorates significantly for passengers.

To address this issue, KT and Samsung analyzed massive amounts of telecommunication quality data collected from high-speed rail corridors, including signal strength and connection failure histories. Traditional maintenance methods typically relied on on-site measurements to address isolated underperforming base stations. In contrast, the new approach leverages large-scale data analytics to identify recurring connection patterns and specific zones prone to service degradation.

The analysis revealed that delayed transitions between consecutive base stations served as the primary cause of communication drops inside train cabins. By pinpointing these troublesome zones, the companies adjusted the handover parameters so that connections to the next base station occur earlier than before.

The field test took place along the Gyeongbu Line between Godeok-myeon in Pyeongtaek and Maesong-myeon in Hwaseong. By comparing telecommunication quality data gathered before and after applying the technology, the partners confirmed that the rate of communication drops during transit dropped by roughly 29 percent, meaning a baseline of 100 connection drops would be reduced to about 71.

Building on these promising results, KT plans to explore options for commercially deploying the technology across high-speed rail networks. Both corporations intend to continue their collaboration in the field of AI-RAN (AI Radio Access Network) to further enhance network performance and operational efficiency.

Jin-guk Jung, Executive Vice President and Head of the Next-Generation Communication Research Center at Samsung Research, stated that the project has proven the capability to enhance the perceived communication quality for customers even in high-speed mobile environments with poor signal conditions.

Jong-shik Lee, Senior Vice President and Head of the Future Network Lab at KT's Network Division, emphasized that the initiative confirmed the potential to improve user experience in vulnerable communication areas like high-speed railways, noting that the companies will persist in developing practical AI-RAN technologies to secure tangible quality upgrades.