基于机器学习的OTFS系统信道估计与信号检测研究进展  

Advancements in Machine Learning-Based Channel Estimation and SignalDetection for OTFS Systems

作　　者：廖勇[1] 韩小金 LIAO Yong;HAN Xiaojin(School of Microelectronics and Communication Engineering,Chongqing University,Chongqing 400044,China)

机构地区：[1]重庆大学微电子与通信工程学院,重庆400044

出　　处：《移动通信》2024年第7期46-56,共11页Mobile Communications

基　　金：重庆市自然科学基金“面向超高速移动场景的OTFS系统信道估计与均衡研究”(CSTB2023NSCQ-MSX0025)。

摘　　要：OTFS调制可以在高速移动场景下有效对抗ICI,能提供比传统的OFDM更显著的性能增益,充分提升高多普勒扩展场景下的通信系统频谱效率,改善通信系统质量,近年来得到广泛研究。OTFS系统的研究重点和难点在于信道估计与信号检测。人工智能已经成为一种有效的信息处理手段,在各行业都得到了广泛的应用。而机器学习是人工智能中的重要分支,通过将机器学习与OTFS技术相结合,可以有效解决信道估计与信号检测中的问题,提高系统性能、稳定性和可靠性。对目前基于机器学习的OTFS系统的信道估计与信号检测算法进行了较为全面的调研、分析、对比和归纳,进而梳理出技术挑战,并就未来的技术发展趋势进行了探讨。OTFS modulation effectively combats ICI in high-mobility scenarios, offering significant performance gains over traditionalOFDM and enhancing spectral efficiency and communication quality in high-Doppler environments, thus attracting extensiveresearch interest in recent years. The primary focus and challenge in OTFS systems lie in channel estimation and signaldetection. Artificial intelligence has emerged as a potent information processing tool, widely adopted across various industries.Machine learning, as a crucial branch of artificial intelligence, when integrated with the OTFS technique, addresses issues inchannel estimation and signal detection, thereby improving system performance, stability, and reliability. This paper provides acomprehensive survey, analysis, comparison, and synthesis of current machine learning-based algorithms for channel estimationand signal detection in OTFS systems, identifies technical challenges, and discusses future technological development trends.

关 键 词：人工智能 机器学习 第六代移动通信 正交时频空 信道估计 信号检测 

分 类 号：TN929.5[电子电信—通信与信息系统]