KT and Nokia Verify AI-Beam Pattern Optimization for 5G

SEOUL, South Korea — KT Corp (KRX: 030200) said Monday it has completed laboratory verification of AI-based beam pattern optimization technology developed in collaboration with Nokia.

The technology applies reinforcement learning algorithms to analyze real-time network data, including traffic flow, and reconfigure beam patterns at 5G base stations based on regional and environmental conditions.

In massive multiple-input multiple-output (mMIMO) deployments—where antennas serve multiple users simultaneously with distinct beams—the number of possible beam pattern combinations within a single cell presents a selection challenge.

The system is designed to evaluate those combinations, select optimal configurations, and adjust strategies through periodic learning based on time of day or environmental variables.

According to KT, the verification results showed improvements in wireless signal quality, coverage stability, and capacity performance within the same hardware environment.

The figures and methodology were not independently verified.

The verification was conducted in February 2026 by KT and Nokia's global research organizations.

KT framed the initiative as part of its AI-RAN strategy, which aims to improve the operational efficiency of the radio access network.

KT and Nokia said they plan to proceed with field verification on commercial networks in the second half of 2026 and indicated they will continue joint work on AI-RAN technologies in connection with 6G standardization efforts.