厦门大学学报（自然科学版）2017，Vol.56Issue(3)：416-422,7.DOI:10.6043/j.issn.0438-0479.201606011
水下冲激噪声环境下基于多特征融合的信号调制方式识别
Classification of Signal Modulation Types Based on Multi-features Fusion in Impulse Noise Underwater
赵自璐 1王世练 1张炜 1谢阳1
作者信息
- 1. 国防科学技术大学电子科学与工程学院,湖南长沙410073
- 折叠
摘要
Abstract
Classification of signal modulation types plays an important role in underwater communication systems.However,present classical classification methods under white Gaussian noises exhibit poor performance under underwater impulse noises.This study proposes a method of modulation classification based on multi-features fusion under underwater impulse noises.First,we apply the nonlinear transformation of the exponential function to eliminating part of impulse noises which are modeled by Alpha stable distribution model.Next,extract the signals' frequency domain features of box dimension,envelopes features of sample entropy and Stockwell transform domain features of energy entropy,and then construct multi-features vectors,which are given to support vector machine (SVM) for fusion and classification.Different signal sets are considered in MATLAB simulating experiments,which provide different number of signal schemes.Results show that the proposed method yields good performance and robustness under impulse noise.In comparison with the simulation,it is verified that the pretreatment of nonlinear transformation significantly improves the performance,and that the multi-features method offers better performance than those single ones do,at the same time increases the identifiable signal types.关键词
冲激噪声/调制识别/Alpha稳定分布/多特征Key words
impulse noise/modulation classification/Alpha stable distribution/multi-features分类
信息技术与安全科学引用本文复制引用
赵自璐,王世练,张炜,谢阳..水下冲激噪声环境下基于多特征融合的信号调制方式识别[J].厦门大学学报（自然科学版）,2017,56(3):416-422,7.
