In this paper, we experimentally demonstrate for the first time an on off keying modulated visible light communications system achieving 170 Mb/s using an artificial neural network (ANN) based equalizer. Adaptive decision feedback (DF) and linear equalizers are also implemented and the system performances are measured using both real time (TI TMS320C6713 digital signal processing board) and offline (MATLAB) implementation of the equalizers. The performance of each equalizer is analyzed in this paper using a low bandwidth (4.5 MHz) light emitting diode (LED) as the transmitter and a large bandwidth (150 MHz) PIN photodetector as the receiver. The achievable data rates using the white spectrum are 170, 90, 40 and 20 Mb/s for ANN, DF, linear and unequalized topologies, respectively. Using a blue filter to isolate the fast blue component of the LED (at the cost of the power contribution of the yellowish wavelengths) is a popular method of improving the data rate. We further demonstrate that it is possible to sustain higher data rates from the white light with ANN equalization than the blue component due to the high signal-to-noise ratio that is obtained from retaining the yellowish wavelengths. Using the blue component we could achieve data rates of 150, 130, 90 and 70 Mb/s for the same equalizers, respectively.
© 2014 Crown
Paul Anthony Haigh, Zabih Ghassemlooy, Sujan Rajbhandari, Ioannis Papakonstantinou, and Wasiu Popoola, "Visible Light Communications: 170 Mb/s Using an Artificial Neural Network Equalizer in a Low Bandwidth White Light Configuration," J. Lightwave Technol. 32, 1807-1813 (2014)