基于CGDNN的低信噪比自动调制识别方法

Low signal-to-noise ratio automatic modulationrecognition method based on CGDNN

针对非协作通信环境中,自动调制识别(automatic modulation recognition,AMR)在低信噪比下泛化能力有限、分类精度不高的问题,提出一种由卷积神经网络、门控循环单元和深度神经网络组成的模型—CGDNN(convolutional gated recurrent units deep neural networks)。首先对I/Q采样信号进行小波阈值去噪,降低噪声对信号调制识别的影响;然后用CNN和GRU提取信号空间和时间特征;最后,通过全连接层进行识别分类。与其他模型对比,验证CGDNN模型在提高AMR性能的同时,显著降低了计算复杂度。实验结果显示,CGDNN模型在RML2016.10b数据集上的平均识别准确率达到了64.32%,提高了-12 dB~0 dB的信号分类精度,该模型大幅减少了16QAM与64QAM的混淆程度,在18 dB时达到了93.9%的最高识别准确率。CGDNN模型既提高了低信噪比下AMR的识别准确率,也提高了模型训练的效率。

To overcome AMR’s limited generalization and low classification accuracy in non-cooperative communication contexts with low signal-to-noise ratio,this paper proposed a model named CGDNN,which integrated CNN,GRU and deep neural networks.To mitigate noise impact on modulation detection,this paper initially denoised I/Q sampled signal using wavelet thresholding.Subsequently,this paper utilized CNN and GRU for extracting spatial and temporal features from signals before proceeding to classification through fully connected layers.Besides enhancing AMR performance,the CGDNN model significantly reduced computational complexity compared to competitors.Experiment results demonstrate an average recognition accuracy of 64.32%on the RML2016.10b dataset,with an enhancement in signal classification accuracy from-12 dB to 0 dB.Moreover,the model substantially decreased confusion between 16QAM and 64QAM,achieving a peak recognition accuracy of 93.9%at 18 dB.CGDNN model effectively improved AMR detection accuracy under low signal-to-noise ratio conditions and enhanced model training efficiency.