KT, Samsung report 29% fewer high-speed rail dropouts in trial

KT and Samsung Electronics reported a 29% reduction in communication interruptions during a trial on South Korea’s Gyeongbu Line high-speed rail, lowering an illustrative interruption index from 100 to about 71.

The demonstration covered the line between Godeok-myeon in Pyeongtaek and Maesong-myeon in Hwaseong. KT said on the 5th that it had developed the radio access network optimization technology with Samsung Electronics; the report does not identify the month.

Neither company disclosed the route length, train frequency, number of base stations or number of users in the test, so the result cannot yet be converted into a network-wide interruption rate.

Samsung Research Executive Vice President Jeong Jin-guk, head of the Next Generation Communication Research Center, said: "We proved that we can improve customers’ perceived communication quality in high-mobility environments."

What problem is the technology addressing?

At 250 km/h or faster, base-station signals change more quickly than in typical communications environments. Consolidated base-station switches therefore occur frequently as trains move between coverage areas.

• A mistimed switch can briefly delay data transmission or reduce call quality.
• Growing use of onboard video increases demand for stable communications during the trip.

Previously, KT responded to poor quality through field measurements focused on specific base stations. KT and Samsung then analyzed signal-strength data collected on high-speed rail sections and records of consolidation failures, using AI and large-scale data to identify recurring problem sections and consolidation patterns.

The companies selected sections where quality repeatedly declined and demonstrated the technology there. That method shifts the operational focus from isolated field checks to patterns visible across the dataset.

What changes in network operation?

After selecting recurring-problem sections, the technology adjusts switching timing so consolidation between base stations occurs earlier. The aim is to complete the switch before quality deteriorates, reducing the interruptions passengers experience while using communications services on board.

The reported result concerns radio-service continuity, not rail capacity. The announcement reports no additional train paths, seats or line throughput, and it gives no capital or operating-cost saving.

For passengers, fewer interruptions mean more reliable calls and data sessions. For KT’s network team, the immediate operating change concerns switch timing rather than train operation or passenger capacity.

How was the improvement tested?

KT and Samsung compared communication-quality data from before and after the technology was applied on the selected sections. The reported rate of interruptions fell by about 29%. On the source’s illustrative scale, 100 interruptions before the change became about 71 afterward.

KT Network Division Senior Vice President Lee Jong-sik, head of the Future Network Lab, said: "This demonstration is a meaningful achievement that confirmed the potential to improve customers’ perceived quality in vulnerable areas such as high-speed rail by using AI."

What remains unproven for wider use?

The 29% figure is a company-reported, section-level result, not a network-wide forecast. The report does not disclose the absolute number of interruptions, observation period, train speeds, train movements, number of devices or number of base stations included.

It also leaves out several items needed to assess wider deployment:

• Separate results for voice calls, data sessions and onboard video.
• A rollout timetable, list of additional sites or procurement plan.
• Capital cost, operating cost, expected payback or effect on network capacity.
• The identity of the rail-service operator, telecom regulator, safety authority or funding body.

No fleet or service timetable accompanies the trial, so the announcement does not show how many trains or passengers on other routes would receive the adjusted network. For KT, the case for wider deployment still rests on repeat measurements across more high-speed sections and under different traffic conditions.

The next useful evidence is a published baseline and absolute post-change count, followed by a rollout plan tied to specific sections and costs. If KT and Samsung reproduce the reduction over a larger sample, the Gyeongbu result can support wider deployment in South Korea; until then, it remains a measured improvement on the selected test sections.