Channel Equalization for OFDM System Based on the BP Neural Network

To improve the performance of the equalizer in OFDM system when the channel has zeros on the FFT grid which may cause the channel matrix uninvertible, this paper introduces a novel equalizer based on the complex-domain back propagation (BP) algorithm. Firstly, the system model and mathematical expression of the OFDM is given in this paper. Secondly, the structure and learning algorithm of complex-domain multilayer perceptron and the equalizer based on BP network are derived. Theoretical analysis and simulation results show that the proposed equalizer based on BP network has better performance than the conventional equalizer