Skip to Main Content
Modulation Classification is an important research subject in the process of designing receiver when signals are transmitted in incooperative communication system. So far, lots of methods of modulation classification for digital signals are limited by the computational complexity, huge calculation or low classification accuracy at low signal noise ratio. In the paper, parameters used for designing classifier are extracted by analyzing the relationship between wavelet transform amplitudes of modulated signals with and without normalization. The algorithms of classifier are simple and the calculation quantity is small. Simulation results show that the anti-noise performance of the algorithm is perfect (success rates are over 97% when Gauss noises in the channel with signal noise ratio not lower than 5 dB).
Wavelet Analysis and Pattern Recognition, 2007. ICWAPR '07. International Conference on (Volume:3 )
Date of Conference: 2-4 Nov. 2007