Dynamic Wavelet Neural Network Feedback Blind Equalization Algorithm based on Fuzzy Control

Aiming at the time-varying and uncertainty of the wireless communications systems, dynamic wavelet neural network feedback blind equalization algorithm based on Fuzzy control was proposed. In the proposed algorithm, the wavelet neural network has dynamic characteristic, by introducing feedback in wavelet neural network and dynamic controlling the variable step-sizes of feedback coefficient and displacement factor, scale factor of the wavelet function in network nodes. For applying complex systems, the input signal and the weights of the network are divided into real and imaginary parts. Compared with traditional neural network equalizer for dynamic system equalization, the proposed algorithm has a better equalization performance in terms of accuracy and stability because of using the current data and historical data at the same time