Automatic Recognition of Analog Modulated Signals Using Artificial Neural Networks

作     者：Jide Julius Popoola Rex Van Olst 

作者机构：Centre for Telecommunications Access and Services School of Electrical and Information Engineering University of theWitwatersrand Johannesburg 2050 South Africa 

出 版 物：《Computer Technology and Application》 (计算机技术与应用（英文版）)

年 卷 期：2011年第2卷第1期

页      面：29-35页

学科分类：11[军事学] 12[管理学] 0810[工学-信息与通信工程] 1105[军事学-军队指挥学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 081002[工学-信号与信息处理] 110503[军事学-军事通信学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

主　　题：Automatic modulation recognition modulation schemes features extraction key artificial neural network (ANN). 

摘      要：This paper presents work on modulated signal recognition using an artificial neural network （ANN） developed using the Python programme language. The study is basically on the analysis of analog modulated signals. Four of the best-known analog modulation types are considered namely： amplitude modulation （AM）, double sideband （DSB） modulation, single sideband （SSB） modulation and frequency modulation （FM）. Computer simulations of the four modulated signals are carried out using MATLAB. MATLAB code is used in simulating the analog signals as well as the power spectral density of each of the analog modulated signals. In achieving an accurate classification of each of the modulated signals, extensive simulations are performed for the training of the artificial neural network. The results of the study show accurate and correct performance of the developed automatic modulation recognition with average success rate above 99.5%.