한국통신학회 KICS

Special Issue on Artificial Intelligence and Machine Learning

Approaches to Communication

Artificial intelligence and machine learning (AI/ML) have achieved remarkable successes in image recognition, voice recognition, natural language processing, robotic control in recent years since the deep learning revolution. Application of AI/ML is now pervasive into many other fields beyond the aforementioned conventional computer science areas, and communication is one of such fields that may benefit from recent advances in AI/ML. The classical framework in communication has evolved into model-based approaches and this is partly owing to excellent modelling of thermal noise as Gaussian noise. In the classical model-based approach to communication, the channel is modeled as a conditional probability for each channel input and the unknown information at the transmitter is then decoded by optimal maximum a posteriori decision rule. In this regard, the development of efficient methods for channel identification and receiver decoding algorithms in addition to capacity-increasing transceiver schemes was the main topic of communication. However, as the frequency of communication increases from mmWave bands to THz and the scale of communication systems reaches the unprecedented level, as envisioned in 6G, the precise modeling of channels is not easy and the conventional design and optimization approaches to communication system design based on concrete analytic models may not be as effective as before. This imposes new challenges to communication system designers and researchers and necessitates rethinking classical communication systems and network design approaches.

The goal of this special issue is to provide a forum for discussing new ideas and recent developments in AI/ML approaches to communication addressing the challenges in current and emerging communication systems and networks. Prospective authors are invited to submit original manuscripts on topics including, but not limited to:

• AI/ML approaches to waveform design and channel coding
• AI/ML approaches to channel estimation and equalization
• AI/ML approaches to interference management
• AI/ML approaches to compensate transceiver nonlinearity and impairments
• AI/ML approaches to MIMO communication
• AI/ML approaches to relay and feedback communication
• AI/ML-based optimization of communication systems and networks
• AI/ML approaches to cross-layer design
• Distributed and federated learning for communications and networking
• Reinforcement learning approaches to MAC, routing, and resource allocation
• Multi-agent reinforcement learning approaches to distributed computing systems
• AI/ML approaches for coexistence and joint communication/sensing
• Quantum AI/ML approaches and applications to wireless communications
• Experimental testbeds for AI/ML-enabled communication systems  

Guest Editors:
Youngchul Sung, KAIST, South Korea, ycsung@kaist.ac.kr
Y.-W. Peter Hong, National Tsing Hua University, Taiwan, ywhong@ee.nthu.edu.tw
Seung Jun Kim, University of Maryland, Baltimore County, USA, sjkim@umbc.edu
Joongheon Kim, Korea University, South Korea, joongheon@korea.ac.kr
Wonjae Shin, Ajou University, South Korea, wjshin@ajou.ac.kr

Important Dates (Tentative):
October 1, 2021: Paper submission deadline
December 1, 2021: Reviews returned to authors
January 1, 2022: Final revised manuscript due
January 31, 2022: Final decision due
March 31, 2022: Publication date

Submission Procedure: