A Novel Adaptive Channel Equalizer Based on Artificial Neural Network Trained by Modified FOA

This paper addresses the problems in channel equalization of high data rates scenarios. Traditional equalization algorithms tend to fall into local optimum, and the complexity is usually high, requiring a long time to converge, thus not suitable for high data rate scenarios. This paper proposes a low complexity and efficient equalizer using artificial neural network (ANN) trained by modified fruit fly optimization algorithm (FOA). After that, we applied it to four different channels and compared it with the ANN trained with back propagation (BP) and ANN trained with particle swarm optimization (PSO) algorithms under the same conditions. It is obvious from the simulation results that the proposed algorithm outperforms the other two algorithms. The algorithm proposed in this paper requires a shorter convergence time and can converge to a smaller value, and the complexity of the algorithm is very low.