Signal Detection and Adaptive Equalization Using a Discrete Wavelet Transform - Artificial Neural Network for OOK Indoor Optical Wireless Links

The paper presents an alternative approach for signal detection and equalization using the discrete wavelet transform (DWT) and the artificial neural network (ANN) in diffuse indoor optical wireless links. The received noisy signal is sampled at multiple rates and the DWT is used for feature extraction while the feedforward backpropagation ANN is used for the pattern classification (i.e. equalization). Different window lengths are investigated and a window of 3-bit is being used for signal classification (i.e. equalization). The simulation results show that for a highly dispersive channel the proposed receiver/equalizer outperforms the traditional equalizer by offering SNR gain of more than 6 dB at the data rate of 155 Mpbs.