| With the wide development of communication technology, the digital modulation mode of the communication signals has become more and more various and the wireless communication environment is becoming more and more complex. In a complex electromagnetic environment in order to accurately demodulate the intercepted signal and then get the useful information, the signal modulation must be recognized. This paper mainly discusses the recognition of digital modulation signals based on Morlet continuous wavelet transform (CWT).This thesis expounds the problem of optimal scales in the process of feature extraction. In order to solve the problem, the modulation recognition can be divided into two parts:the modulation recognition of ten kinds of single carrier signals and the modulation recognition between OFDM(Orthogonal Frequency Division Multiplexing) and single carrier signals. The main works and achievements in this thesis are as follows:1)The thesis proposes an algorithm based on optimized-scale continuous wavelet transform. For each feature that is used in recognition, the corresponding optimal scale is calculated. Finally a set of recognition system is designed with the optimal scales and decision tree classifier which has the biggest ability to resist noise. The simulation result indicates that an overall success rate of over90%can be reached when the SNR (Signal) is not less than2dB.The good results came from the use of optimal scale.2) The wavelet scales that is used two times in the CWT are discussed. The method of control variables is used in the study of the influence that the scales have on signal feature. When the first scale is3 and the second scale is25, the feature of OFDM has the biggest difference from the single carrier signals’. The optimal scales can make the feature have the greatest ability to resist noise. The simulation result indicates that when the SNR is greater than1dB, the success rate of recognition of modulation rate of OFDM and single carrier signals can reach100%. |