首页|ResNet-UAN-AUD:基于深度学习的水声上行非正交多址通信系统活动用户检测方法

ResNet-UAN-AUD:基于深度学习的水声上行非正交多址通信系统活动用户检测方法

扫码查看
原文链接
  • 国家科技期刊平台
  • NETL
  • NSTL
  • 万方数据
水下声学网络(Underwater Acoustic Networks,UAN)是探测未知水域的重要技术手段.非正交多址(Non-Orthogonal Multiple Access,NOMA)是一种新颖的移动通信技术,支持时域、频域、空域/编域的非正交分配,可有效地提高网络容量和用户接入数,为性能和电量受限的UAN提供创新解决方案.活动用户检测(Active User Detection,AUD)是NOMA通信系统的基础支撑,对于NOMA系统消除信号干扰和提高接收性能至关重要.ResNet是基于残差模块跳跃连接的神经网络,解决了深度学习的梯度消失和网络退化问题.提出了一种基于深度学习的水声上行NOMA通信系统AUD检测方案.首先,构建水声上行NOMA通信系统基本模型;其次,实施NOMA活动用户检测问题的数学表征;接着,开发基于ResNet网络的水声NOMA系统活动节点检测方法(ResNet-UAN-AUD);最后,执行仿真实验.结果表明,ResNet-UAN-AUD的检测性能接近基于长短期记忆网络的活动用户检测(LSTM-UAN-AUD)方案,而复杂度略高于基于卷积神经网络的活动用户检测(CNN-UAN-AUD)技术,实现了次优目标,适合水声上行NOMA系统使用.
ResNet-UAN-AUD:An Active User Detection Method for Underwater Acoustic Uplink NOMA Communication System Based on Deep Learning
Underwater acoustic networks(UANs)are the primary technical means of detecting unknown waters.Non-orthogonal multiple access(NOMA)is a novel communications technology that supports non-orthogonal resource allocation in the time,frequency,or space/code domains,which can effectively improve network capacity and user access,providing innovative solutions for performance and power-constrained UANs.Active user detection(AUD)is essential for the NOMA system to eliminate signal interference and improve re-ception performance.ResNet is a neural network based on residual module hopping connection,which solves the problem of gradient disappearance and network degradation in deep learning.A ResNet-based AUD detection scheme(ResNet-UAN-AUD)is proposed for a hydroacoustic uplink NOMA system.Firstly,the basic model of the hydroacoustic uplink NOMA network is established.Secondly,the mathematical characterization of the AUD problem is realised.Thirdly,the ResNet-UAN-AUD is developed.Finally,the experimental simulation of the proposed scheme is carried out.The results show that the performance of ResNet-UAN-AUD is close to that of the ac-tive user detection scheme based on the long short-term memory network(LSTM-UAN-AUD).The complexity is slightly higher than that of the active user detection method based on the convolutional neural network(CNN-UAN-AUD),which achieves the suboptimal objec-tive and fits the hydroacoustic uplink NOMA system.

underwater acoustic networkdeep learningresidual neural network(ResNet)active user detectionuplink NOMA commu-nication system

王建平、陈光岚、冯启高、马建伟

展开 >

河南科技学院信息工程学院,河南 新乡 453003

河南科技大学信息工程学院,河南 洛阳 471000

水声网络 深度学习 残差神经网络(ResNet) 活动用户检测 上行NOMA通信系统

河南省科技计划项目河南省科技计划项目河南省科技计划项目河南省高等学校青年骨干教师计划项目河南省高等学校重点科研计划项目2021年度国家级大学生创新训练项目重点支持领域项目2021年度新乡市重大专项河南省重点研发专项

2321021111282221023201812221021100112019GGJS17223B52000320211046700121ZD003241111211800

2024

10.3969/j.issn.1004-1699.2024.06.008

传感技术学报
东南大学 中国微米纳米技术学会

传感技术学报

CSTPCD北大核心
影响因子：1.276
ISSN：1004-1699
年,卷(期)：2024.37(6)
  • 45