无线电工程2025，Vol.55Issue(2)：291-297,7.DOI:10.3969/j.issn.1003-3106.2025.02.008
基于卷积自适应降噪网络的自动调制识别方法
Automatic Modulation Recognition Method Based on Convolutional Adaptive Denoising Network
陈昊 1郭文普 1康凯 1施昊1
作者信息
- 1. 火箭军工程大学 作战保障学院,陕西 西安 710025
- 折叠
摘要
Abstract
To deal with the problem of low recognition accuracy of Automatic Modulation Recognition(AMR)methods under low Signal to Noise Ratio(SNR)conditions,an AMR method is proposed based on convolutional Adaptive Noise Reduction(ANR).In this method,phase transformation is used to reduce the impact of phase shift on modulation recognition;Convolutional Neural Network(CNN)and Gated Recurrent Unit(GRU)are used to extract spatial and temporal features of signals,respectively;an ANR Module is added after CNN to perform adaptive soft thresholding on convolutional features under different SNR conditions to improve network robustness.The simulation results on the benchmark dataset RML2016.10a show that the proposed model achieves better recognition accuracy compared to other network models when the SNR is greater than-8 dB.关键词
自动调制识别/卷积神经网络/自适应降噪模块/软阈值Key words
AMR/CNN/ANR module/soft thresholding分类
信息技术与安全科学引用本文复制引用
陈昊,郭文普,康凯,施昊..基于卷积自适应降噪网络的自动调制识别方法[J].无线电工程,2025,55(2):291-297,7.
