Effective denoising and adaptive equalization of indoor optical wireless channel with artificial light using the discrete wavelet transform and artificial neural network

Indoor diffuse optical wireless (OW) communication systems performance is limited due to a number of effects; interference from natural and artificial light sources and multipath induced intersymbol interference (ISI). Artificial light interference (ALI) is a periodic signal with a spectrum profile extending up to the MHz range. It is the dominant source of performance degradation at low data rates, which can be removed using a high-pass filter (HPF). On the other hand, ISI is more severe at high data rates and an equalizing filter is incorporated at the receiver to compensate for the ISI. This paper provides the simulation results for a discrete wavelet transform (DWT)-artificial neural network (ANN)-based receiver architecture for on-and-off keying (OOK) non-return-to-zero (NRZ) scheme for a diffuse indoor OW link in the presence of ALI and ISI. ANN is adopted for classification acting as an efficient equalizer compared to the traditional equalizers. The ALI is effectively reduced by proper selection of the DWT coefficients resulting in improved receiver performance compared to the digital HPF. The simulated bit error rate (BER) performance of proposed DWT-ANN receiver structure for a diffuse indoor OW link operating at a data range of 10-200 Mbps is presented and discussed. The results are compared with performance of a diffuse link with an HPF-equalizer, ALI with/without filtering, and a line-of-sight (LOS) without filtering. We show that the DWT-ANN display a lower power requirement when compared to the receiver with an HPF-equalizer over a full range of delay spread in presence of ALI. However, as expected compared to the ideal LOS link the power penalty is higher reaching to 6 dB at 200 Mbps data rate.