深度学习在基于正交频分复用技术中的应用研究  

Research on application of deep learning in orthogonal frequency division multiplexing technology

作　　者：马洪灿 李良彬 MA Hongcan;LI Liangbin(Yuncheng County Public Security Bureau,Heze,Shandong 274700,China)

机构地区：[1]郓城县公安局,山东菏泽274700

出　　处：《计算机应用文摘》2025年第9期64-66,共3页

摘　　要：深度学习是一种强大的拟合器,在通信领域中具有广泛的应用价值。作为通信系统的根基之一,正交频分复用(Orthogonal Frequency Division Multiplexing,OFDM)存在耗费频谱资源较多、子载波间干扰严重等不足,导致OFDM系统恢复失真信号的能力有所下降。针对此问题,探究深度学习在OFDM技术中的应用具有重要意义。研究结果显示,深度学习为OFDM技术提供了高效的解决方案。相较于现有技术,基于深度学习的OFDM无线通信技术、水声通信技术、索引调制策略和自编码器方案有着明显的优势。在导频和循环前缀较少的前提下,其可达到更优的误码率性能,并维持低复杂度。Deep learning is a powerful fit and has a wide range of applications in the field of communication.As one of the foundations of communication systems,Orthogonal Frequency Division Multiplexing(OFDM)consumes a lot of spectrum resources,serious interference between subcarriers and other shortcomings,resulting in the OFDM system̓s ability to recover distorted signals decreased.To solve this problem,it is of great significance to explore the application of deep learning in OFDM technology.The results show that deep learning provides an efficient solution for OFDM technology.Compared with the existing technology,OFDM wireless communication technology,underwater acoustic communication technology,index modulation strategy and autoencoder scheme based on deep learning have obvious advantages.Under the premise of fewer pilot and loop prefixes,it can achieve better bit error rate performance and maintain low complexity.

关 键 词：OFDM 通信技术 索引调制 自编码器 深度学习 

分 类 号：TN92[电子电信—通信与信息系统]