基于被动时间反转-自编码器的水声通信信号调制识别方法  被引量:7

Modulation recognition method of underwater acoustic communication signal based on passive time reversal-autoencoder

在线阅读下载全文

作  者:胡雅琳 付晓梅 暴纪欣 闫红刚[2] 孙万忠 钱治文 HU Yalin;FU Xiaomei;BAO Jixin;YAN Honggang;SUN Wanzhong;QIAN Zhiwen(School of Marine Science and Technology,Tianjin University,Tianjin 300072,China;School of Information System Engineering,Information Engineering University,Zhengzhou 450001,China)

机构地区:[1]天津大学海洋科学与技术学院,天津300072 [2]信息工程大学信息系统工程学院,河南郑州450001

出  处:《厦门大学学报(自然科学版)》2023年第4期590-597,共8页Journal of Xiamen University:Natural Science

基  金:国家自然科学基金(62171310)。

摘  要:海洋环境中存在的多径效应会影响水声通信信号的特征,降低调制识别准确率.为了减小多径效应对信号调制识别造成的干扰,本文提出了一种基于被动时间反转-自编码器的信号增强方法.利用水声通信信号中常见的同步信号作为探测信号,实现无外加信号的被动时间反转以及功率谱和平方谱两种频域特征的增强,并设计了卷积神经网络进行调制识别.网络模型经过Bellhop仿真水声信道和少量非测试环境的实测数据进行训练后,在3种与训练环境不同的实际水声信道下测试,所有调制信号识别准确率均高于80%,6种调制信号的平均识别准确率与已有文献中的方法相比至少提高了20个百分点.The multipath effect in the marine environment affects characteristics of underwater acoustic communication signals and reduces the accuracy of modulation recognition.To reduce the interference caused by this multipath effect on signal modulation recognition,we propose a signal enhancement method based on passive time reversal-autoencoder.Common synchronous signals in underwater acoustic communication signals are used as detection ones to attain passive time reversal without external signals and the enhancement of two frequency domain features of power spectrum and square spectrum.A convolutional neural network is designed for modulation recognition.After the network model is trained by Bellhop simulated underwater acoustic channel and a small amount of measured data in non-tested environment,it is tested in three actual underwater acoustic channels that differ from the training environment.The recognition rate of all modulation signals exceeds 80%,and the average recognition accuracy of six modulation signals is improved by at least 20 percentage points compared with methods reported in the literature.

关 键 词:调制识别 多径效应 被动时间反转-自编码器 卷积神经网络 

分 类 号:TN911.3[电子电信—通信与信息系统]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象