Department of Electronic Engineering Digital Transmission Research Laboratory Develops 6G Smart Relay AI Beam Management Technology through Joint Research with ETRI
Department of Electronic Engineering Digital Transmission Research Lab, ETRI collaborated on the development of 6G smart relay AI beam management technology
▲ (from left) Hyunseok Noh Ph.D. program, Electronics and Telecommunications Research Institute (ETRI) logo
Department of Electronic Engineering Digital Transmission Research Lab(advisor Wonjin Seong)has developed 'AI-based Smart Relay Beam Management Technology' through joint research with the Electronics and Telecommunications Research Institute (ETRI) to address blind spots in next-generation mobile communication networks and improve transmission performance. This research was conducted as part of the Ministry of Science and ICT's national project, 「6G Communication Coverage Expansion Relay System Technology Development」, and was carried out in collaboration with major domestic academic and industrial institutions, including Sogang University and ETRI, as well as LG Electronics, KT, SK Telecom, and LG Uplus, to verify the practicality of the technology. The research was supported by the Ministry of Science and ICT, and the results of the study have contributed to the development of 6G communication technology, and the participating institutions have validated the technology's practicality. Through this research, the Department of Electronic Engineering Digital Transmission Research Lab has made significant contributions to the field of electronic engineering, and the lab's research results are expected to have a major impact on the development of future communication technologies, with the participation of major domestic institutions, including LG Electronics, KT, SK Telecom, and LG Uplus.
The research team established a 3dimensional digital twin-based mobile communication simulation environment that accurately reflects the building structure and radio wave characteristics of actual urban areas, and proposed a method to greatly improve the transmission performance of the network control relay(NCR) using this environment. In particular, the team overcame the limitations of using existing fixed beam sets, and developed an 'AI-based adaptive beam management technology' that enables the relay to learn and update the optimal beam set based on user distribution and signal measurement information·on its own, which was independently developedby the team, and verified through application tests in various actual urban environments, such as Seoul and Boston in the US, that it can achieve significant transmission efficiency improvement without additional base station-relay signal overhead or standard changes compared to existing methods–and confirmed that this improvement can be achieved.
[Major Academic Achievements]
Hongseok Noh, a Ph.D. student in the Department of Electronic Engineering who led this research, presented the core technology of learning-based beam management at IEEE ICC 2026 (in 5 months of ) the world-renowned international academic conference in the field of communication.
Then, it was developed into a technology based on reinforcement learning and transfer learning considering the hardware constraints of actual relays, and a paper was published in IEEE Transactions on Wireless Communications (TWC, IF 10.3, JCR top 4%) in the October 2026 issue and achieved the result of being published in 10 the top international academic journal in the field of wireless communication.
Our research team's achievement is expected to be utilized as a core element technology for next-generation 6G AI-Native mobile networks through combination with intelligent real-time wireless resource management technology in the future.
▶ Paper Information
1. International Conference Paper (IEEE ICC 2026)
- Paper Title: On-Device AI/ML Codebook Self-Learning for Enhanced Beamforming of Network-Controlled Repeaters
- Conference Name: IEEE International Conference on Communications (ICC 2026, 2026 5month presentation)
- Authors: Hongseok Noh (Corresponding 1author), Wonjin Sung (co-corresponding author)
2. International Journal Papers(IEEE TWC)
- Paper Title: Hardware-Constrained Reinforcement Learning for Adaptive Beam Management in Network-Controlled Repeaters: A Transfer Learning Approach
- Journal title: IEEE Transactions on Wireless Communications (2026 year 10 issue)
- Journal impact: IF 10.3 / JCR Q1 top 4% (based on 2025 JCR standards)
- Author: Hyunseok Noh (et al.1Author), Wonjin Seong (Corresponding author)