摘要
众多学者深入研究了信号调制识别方法,但对于多进制相移键控信号(MPSK)的识别效果不佳。本文提出基于深度学习的识别模型,利用时频图和星座图特征识别MPSK信号类型,并融合两种识别结果进行决策。对比三种方法,有效提升了MPSK信号的识别效果。
Numerous scholars have studied signal modulation recognition methods in depth,but they are not effective in recognising Multi-Progressive Phase Shift Keying(MPSK)signals.This paper proposes a recognition model based on deep learning,which uses time-frequency diagram and constellation diagram features to identify MPSK signal types,and fuses the two recognition results for decision making.Comparing the three methods,the recognition effect of MPSK signals is effectively improved.
出处
《现代传输》
2024年第4期72-75,共4页
Modern Transmission
关键词
深度学习
调制识别
多进制相移键控
决策融合
deep learning
modulation identification
multivariate phase shift keying
decision fusion
