摘要
提出数字调制信号的人工神经网络识别方法，从信号幅度、相位、频率及功率谱等特性中提取五种特征参数，用于训练神经网络对数字调制信号的识别。采用神经网络，不仅可提高识别的智能化，而且能提高正确识别率。
A new algorithms of digital modulation recognition is presented in this article. Artificial neural networks(ANNs) are used to recognize digital modulation in this approach, which is suitable for digital modulation types including 2ASK,4ASK,2FSK,4FSK,2PSK,4PSK. ANNs approach can automatically chose the threshold for each feature. It not only makes the classification intelligible, but also improves the success rate of classification.
出处
《国防科技大学学报》
EI
CAS
CSCD
1999年第2期58-61,共4页
Journal of National University of Defense Technology
关键词
数字调制信号
神经网络
信号识别
特征提取
modulation recognition, digital modulation, artificial neural networks
