Adaptive Turbulence Channel Equalization Based on Neural Network

Because of the advantages of high speed, high capacity and low power consumption, Free-Space Optical (FSO) communication is used in high-speed access networks widely. However, the effects of light intensity flicker will increase the Bit Error Rate (BER) of the recovered signal. Besides, with the increasing of communication rate and limitation of optical devices cost, the problem of Inter Symbol Interference (ISI) becomes more serious. In order to solve the ISI in FSO, We design an adaptive equalizer (RBF-RNNE) cascaded with radial basis function neural network (RBFNN) and recurrent neural network (RNN), The equalizer can overcome ISI and improve the communication quality based on the fast convergence of RBFNN and the convenience of RNN to handle time-series signals. The simulation results show that the equalizer converges faster than the RNN equalizer with the same training data; RBF-RNNE can reach 3.8*10(-3) at a signal-to-noise ratio (SNR) of 12.5dB approximately, which is 1.5dB higher than RNN equalizer and 1.9dB higher than RBF equalizer, and can reduce the BER in turbulent channels effectively.