Professor Jungsuek Oh of the Department of Electrical and Computer Engineering at Seoul National University has been selected for the Technology category of the 2025 Samsung Future Technology Development Program. In recent years, Prof. Oh has led research that integrates AI with radio technology to address major industrial challenges. The excellence of his research group has been widely recognized through distinctions such as the Samsung Electronics DX Division Head IT Innovation Award.

The Samsung Future Technology Development Program is a nonprofit research funding initiative launched by Samsung Electronics in 2013, with a total endowment of 1.5 trillion KRW. The program aims to advance fundamental science, drive innovation in industrial technologies, address pressing societal challenges through science and technology, and cultivate world-class scientific talent in Korea. It supports creative frontier research in fundamental science, the foundation of scientific and technological progress; materials science, which underpins manufacturing; and ICT and convergence technologies, which enable industrial advancement and the creation of new markets.

Prof. Oh will conduct a four-year research and development project titled “Implementation and Performance Verification of an Ultra-Dense 6G MIMO Antenna-DPD Integrated System Based on Probabilistic Inverse-Design AI Models.”

Project Overview:

For next-generation 6G communications, the Upper-Mid Band (7-24 GHz) spectrum is drawing increasing attention. However, this band faces key limitations: lower data rates compared to mmWave 5G and degraded coverage relative to existing Sub-6GHz systems. To address these challenges, active research is underway on ultra-dense MIMO systems, which maximize communication capacity and coverage by packing a greater number of antennas within a given aperture. Yet, in such extremely dense array environments, antennas are positioned so closely that coupling increases sharply, and the limited physical area forces components for isolation, matching and Envelope Correlation Coefficient (ECC) control to be placed in very close proximity. This results in strong interdependence among design variables and lengthy design cycles, since performance improvements rely heavily on structural modifications. To overcome these limitations, the proposed research seeks to develop a probabilistic inverse-design model based on generative AI techniques, such as diffusion models, that can effectively model the vast design space and complex nonlinear relationships inherent to ultra-dense MIMO arrays. Nonetheless, severe coupling and limited area impose unavoidable practical performance limitations, making it difficult to achieve high levels of isolation and matching conditions. These constraints introduce uncertainty in securing optimal performance for MIMO systems integrated with Digital Predistortion (DPD). Moreover, conventional DPD methods—that inadequately account for dynamic operation conditions—struggle to provide optimized performance when applied to ultra-dense antenna arrays. As such, this project aims to establish an AI-driven antenna-DPD design framework to address the practical performance limitations encountered in extreme ultra-dense MIMO environments. Through this framework, the research seeks to develop cost-efficient, high-data-rate MIMO communication systems based on ultra-dense antenna architectures and to explore novel structural design possibilities that conventional approaches have been unable to reach.

Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57047

Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr