摘要
信号调制样式的自动识别是软件无线电必备的功能之一,基于人工神经网络的识别方法因其较其他方法具有更好的性能受到广泛关注。分析了基于神经网络调制信号识别技术的基本原理,将目前研究的调制信号识别分为基于多层感知器神经网络的调制信号识别和基于径向基函数神经网络的调制信号识别,提出了神经网络调制信号识别技术进一步的研究方向。
The auto-recognition of the signal medulation mode is one of the necessary functions of the software based radio（SBR）.The recoghnition method based on the artificial neural network（ANN）arouses widespread interest because of its better performance than that of the other methods.This paper analyzes on the basic principles of the neural network based signal modulation recognition technique,devides the current signal modulation recognition into the signal modulation recogaition based on the multi-level preceptor（MLP）and the signal modulation recognition based on the redial-based funtion（RBF）,and puts forward the diretions of the further researches on the signal modulation recognition technique based on ANN.
出处
《科技情报开发与经济》
2006年第4期237-238,F0004,共3页
Sci-Tech Information Development & Economy
关键词
神经网络
多层感知器
径向基函数神经网络
调制信号识别
artificial neural network（ANN）
multi-level preceptor（MCP）
redial-based function（RBF）neural network
signal modulation recognition
