摘要
随着通信环境的日益复杂,对低信噪比下的调制信号识别要求日益提高。针对低信噪比下信号识别率较低的问题,提出了一种基于信号变换域特征的改进残差神经网络调制识别算法,利用信号在自相关域和离散余弦变换域(Discrete Cosine Transform,DCT)表现的特征不同,将信号特征作为并联网络的输入,在网络内部对提取到的不同特征进行融合,在低信噪比下,对BPSK,QPSK,8PSK,16QAM,32QAM,64QAM,128QAM七类调制信号进行自动识别。实验结果表明,当信噪比为0 dB时,7类信号的平均识别率接近93%;信噪比为1 dB时,每一类信号的识别率均达到90%。
With the increasing complexity of communication environment,the requirements for the recognition of modulated signals under low signal-to-noise ratios have gradually become higher.To address the low signal recognition rate under low signal-to-noise ratio,an improved residual neural network modulation recognition algorithm based on signal transform domain characteristics is proposed.Based on the difference of the characteristics of signal in the autocorrelation domain and the discrete cosine transform domain(Discrete Cosine Transform,DCT),the signal characteristics are used as the input of the parallel network,and the different extracted characteristics are integrated inside the network.Under low signal-to-noise ratio,the characteristics of seven types of modulation signals,i.e.BPSK,QPSK,8 PSK,16 QAM,32 QAM,64 QAM and 128 QAM,are automatically recognized.The experimental results show that when the signal-to-noise ratio is 0 dB,the average recognition rate of the seven types of signals is close to 93%,and when the signal-to-noise ratio is 1 dB,the recognition rate of each type of signal reaches 90%.
作者
廖星
高勇
LIAO Xing;GAO Yong(College of Electronics and Information Engineering,Sichuan University,Chengdu 610065,China)
出处
《无线电工程》
北大核心
2022年第9期1574-1579,共6页
Radio Engineering
关键词
调制识别
自相关
DCT变换
神经网络
modulation recognition
self-correlation
DCT transformation
neural networks
