基于全连接神经网络的载波频率偏移估计方法  

Carrier Frequency Offset Estimation Method Based on Fully Connected Neural Network

作　　者：周兴 ZHOU Xing(Faculty of Electrical Engineering and Computer Science,Ningbo University,Ningbo 315211,China)

机构地区：[1]宁波大学信息科学与工程学院,宁波315211

出　　处：《无线通信技术》2022年第4期6-11,共6页Wireless Communication Technology

摘　　要：在正交频分复用(orthogonal frequency division multiplexing,OFDM)通信系统中,由于发射机和接收机晶体振荡器固有的物理特性不同,导致载波频率偏移(carrier frequency offset,CFO)的产生,它破坏了OFDM信号子载波之间的正交性,影响信号的有效传输。传统的载波频率偏移估计方法是使用训练符号或导频符号来完成,这占用了一定通信带宽并且增加了计算开销。本文提出了一种基于深度学习的载波频率偏移估计方法,使用全连接神经网络,以数据驱动的方式,直接根据接收信号的数据部分估计出载波频率偏移。经过实验仿真可知,所提方法可以提升带宽效率,并且比传统估计方法有更高的估计准确性。In orthogonal frequency division multiplexing(OFDM)communication system,due to the inherent physical characteristics of the transmitter and receiver crystal oscillator are different,resulting in carrier frequency offset(CFO),which destroys the orthogonality between the subcarriers of OFDM signals and affects the effective transmission of signals.The traditional carrier frequency offset estimation method is to use training symbols or pilot symbols to complete,which occupies a certain communication bandwidth and increases the computational overhead.In this paper,a carrier frequency offset estimation method based on deep learning is proposed,which uses fully connected neural network to estimate the carrier frequency offset directly from the data part of the received signal in a data-driven manner.Experimental results show that the proposed method can improve bandwidth efficiency and has higher estimation accuracy than traditional estimation methods.

关 键 词：OFDM 载波频率偏移 深度学习 全连接神经网络 

分 类 号：TN92[电子电信—通信与信息系统]