Prof. Sang-Hyo Kim’s research team at the Department of Electrical and Computer Engineering, Sungkyunkwan University (IITP NRC: SKKU Next-Generation Channel Coding Research Center) has developing next-generation wireless error correction code technologies powered by artificial intelligence (AI), establishing a foundation to lead 6G and future communication technologies.

In this study, Prof. Kim’s team developed a Multiple-Masks Attention–based decoding method built upon the Transformer architecture, a core structure of large language models. By leveraging the structural diversity of codes, this approach significantly improves the decoding performance of short block error correction codes and demonstrates the potential for application in ultra-reliable low-latency communications (URLLC) for autonomous driving, industrial IoT, and AI-based wireless networks (AI-RAN).

In addition, the team applied a boosted learning method to neural decoders for LDPC (Low-Density Parity-Check) codes, which are currently used in 5G communication systems, achieving extremely low error rates. This result meets the ultra-reliability requirements demanded by 6G, marking an important milestone that is expected to contribute to future 6G standardization and commercialization.

These research achievements were realized through collaboration with Prof. Yongjune Kim (POSTECH), Prof. Hee-Youl Kwak (University of Ulsan), Dr. Seong-Joon Park (POSTECH), and Emeritus Prof. Jong-Seon No (Seoul National University). The related technologies were published as two papers in the IEEE Journal on Selected Areas in Communications (JSAC) (JCR Top 1.0%, IF 17.2) in April and July 2025. Furthermore, the team presented their work on the Cross-Message Passing Transformer (CrossMPT) decoder at ICLR 2025, one of the world’s top three conferences in machine learning and deep learning, where the academic and technical value of their AI-based error correction technology was internationally recognized.

Prof. Kim stated, “AI technology is providing a new paradigm for wireless communications. We expect our research to contribute to the advancement of 6G technologies, AI-native networks, machine-to-machine and AI-to-AI communications, and ultimately the realization of semantic communications.”

Established in 2024, the IITP-NRC Next-Generation Channel Coding Research Center at SKKU is the only dedicated research hub in Korea focusing on channel coding (error correction code) technologies for 6G and future communications. The program will continue through 2031.

This research has been supported by the Network Research Center (NRC) program of the Institute for Information & Communications Technology Planning & Evaluation (IITP), Channel Coding/Decoding and Channel Estimation for Next-Generation Communications, and by the National Research Foundation of Korea (NRF).

※ Paper 1: Multiple-Masks Error Correction Code Transformer for Short Block Codes (Published in July 2025)
※ Paper 2: Boosted Neural Decoders: Achieving Extreme Reliability of LDPC Codes for 6G Networks (2025년 4월 게재)
※ Journal: IEEE Journal on Selected Areas in Communications (JCR top 1% in Electrial Engineering)

▲ IIT-NRC SKKU Next Generation Channel Coding Research Center

▲ Architecture of Error Correction Code Transformer with Multiple Masks