Dynamic Recurrent Fuzzy Wavelet Neural Network Blind Equalization Algorithm

摘要

For dynamic system, the traditional fuzzy neural network blind equalization algorithm is bad in equalization performance. In order to overcome this shortcoming, a dynamic recurrent fuzzy wavelet neural network blind equalization algorithm is proposed. While combining the wavelet neural network with static fuzzy neural network and making full use of strong reasoning capacity and powerful adaptability of fuzzy neural network, this proposed algorithm adds memory unit between the normalization layer and the fuzzy layer of the fuzzy neural network, and introduces feedback in the fuzzy neural network, in this way, the proposed algorithm does well in dynamic system. Due to take advantage of strong approximation ability of wavelet function, it also embeds wavelet function into the fuzzy neural network and makes the convergence performance improve greatly. Simulation results show that the convergence performance of the proposed algorithm is better than the general fuzzy neural network blind equalization algorithm.