基于深度迁移学习的无线通信数据自适应分类研究  

Adaptive Classification of Wireless Communication Data Based on Deep Transfer Learning

作　　者：胡龙龙 刘伟康 HU Longlong;LIU Weikang(School of Information and Electronic Engineering,Shangqiu Institute of Technology,Shangqiu 476000,China)

机构地区：[1]商丘工学院信息与电子工程学院,河南商丘476000

出　　处：《通信电源技术》2025年第5期1-3,共3页Telecom Power Technology

摘　　要：由于无线通信数据类别标签标注困难,导致传统依赖于大量标注数据的分类方法的准确性和可靠性受限,研究提出一种基于深度迁移学习的无线通信数据自适应分类。采用小波阈值法去噪,基于深度学习在源域上预训练卷积神经网络模型,迁移至目标域微调。在微调后的卷积神经网络模型中输入去噪后的无线通信数据,实现数据自适应分类。实验结果表明,该方法的AUC值为0.963,较对照组提升0.077、0.094,验证了该方法的有效性和优越性。Due to the difficulty of labeling wireless communication data,the accuracy and reliability of traditional classification methods that rely on a large number of labeled data are limited.This paper proposes an adaptive classification of wireless communication data based on deep transfer learning.The wavelet threshold method is used to denoise,and the convolutional neural network model is pre-trained in the source domain based on deep learning,and then moved to the target domain for fine tuning.Input the denoised wireless communication data into the fine-tuned convolutional neural network model to realize adaptive data classification.The experimental results show that the AUC value of this method is 0.963,which is 0.077 and 0.094 higher than that of the control group,which verifies the effectiveness and superiority of this method.

关 键 词：深度迁移学习 无线通信 通信数据 自适应分类 

分 类 号：TN9[电子电信—信息与通信工程]