摘要
自动调制分类(AMC)在频谱监测和认知无线电中具有重要意义。近年来,Chirp扩频通信(CSS)由于其良好的抗干扰能力和稳健性得到了较大发展,但是对CSS信号的AMC方法却鲜有研究。针对这种情况,该文提出了一种基于多特征融合(MFF)的CSS信号调制分类方法,利用频谱和时频图特征融合学习并引入注意力模块来实现CSS信号调制识别。对11类CSS信号调制样式的仿真实验结果表明,该方法有优越的识别性能。
Automatic Modulation Classification(AMC)is essential for spectrum monitoring and cognitive radio.The Chirp Spread Spectrum(CSS)communication scheme could be developed remarkably due to its good antiinterference ability and robustness.However,research on the AMC of the CSS communication scheme is limited.Therefore,this paper proposes a CSS signal modulation classification method based on Multi-Feature Fusion(MFF)to enhance its recognition accuracy.This method which leverages spectrum and time-frequency map feature fusion learning and incorporates an attention module.The results of 11 types of CSS formats demonstrate that the proposed scheme exhibits superior recognition performance.
作者
王翔
宋川江
杨战鹏
WANG Xiang;SONG Chuanjiang;YANG Zhanpeng(State Key Laboratory of Complex Electromagnetic Environment Effects on Electronics and Information System,College of Electronic Science,Natioanal University of Denfense Technology,Changsha 410073,China;Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100190,China)
出处
《电子与信息学报》
EI
CSCD
北大核心
2023年第11期4003-4015,共13页
Journal of Electronics & Information Technology
基金
国家自然科学基金(62271494)。
关键词
CHIRP信号
CSS信号
自动调制分类
多特征融合
Chirp signal
Chirp Spread Spectrum(CSS)
Automatic Modulation Classification(AMC)
Multi-Feature Fusion(MFF)
