摘要
针对目前多种信号样式及低信噪比下信号识别率低的问题,提出一种基于BP神经网络的复杂信号识别技术。通过分析并提取具有高区分度的信号特征,设计三层BP神经网络,重点研究多种、复杂通信调制信号的识别性能,实现基于BP神经网络的识别算法。仿真验证表明,该算法对信号的适应性较高,对17种复杂信号在5 dB信噪比时达到80%以上的正确识别概率,在7 dB时达到95%以上的正确识别概率。
A signal modulation recognition technology based on BP neural network is proposed,to resolve the low recognition rate under multiple signal types with low Signal-to-Noise-Ratio(SNR).Signal features which can separate different signals are analyzed and extracted,and a three-layer BP neural network is designed,the performance of recognition is researched,and the modulation recognition is realized based on BP neural network.Simulation results show that,the algorithm is well adapted to different signals.For all the 17 signal types,the correct recognition ratio is above 80%at 5 dB SNR,and above 95%at 7 dB SNR.
作者
易云清
吕乐群
卢圆圆
刘敏
YI Yunqing;LV Lequn;LU Yuanyuan;LIU Min(Science and Technology on Electronic Information Control Laboratory,Chengdu 610036,China)
出处
《电子信息对抗技术》
2020年第6期16-21,共6页
Electronic Information Warfare Technology
关键词
调制识别
神经网络
深度学习
信号特征
modulation recognition
neural network
deep learning
signal feature
