Cognitive digital modulation classifier using artificial neural networks for NGNs

An automatic cognitive classification scheme for modulation is developed and presented in this paper. With the help of this scheme, the classification and recognition cognitively of wireless communication signals is possible. The proposed features of the scheme have the possibility to adapt it dynamically to any modulation technique, and the capability to identify it correctly. The developed scheme is based on temporal or statistical parameters have been used to identify M-ary PSK, M-ary QAM, and FSK modulations. The simulated results show that the modulation classification is possible at very low SNR. Experimental results lead us to a recognition percentage of 99%, when SNR is 1 dB.