Chebyshev Polynomial-Based Adaptive Predistorter for Nonlinear LED Compensation in VLC

In recent times, there has been a surge in research on visible light communications (VLC). In VLC systems, normal light emitting diode (LED) lamps are used as transmitters by modulating the optical power of the LED with the input current. All this is done at a high frequency, so that the fluctuations are not visible to the naked eye. It is a well known fact that the LED characteristics are not linear. Such nonlinearities degrade the performance of VLC. In the literature, a predistorter using a linear adaptive scaling parameter is proposed as a predistorter, which uses the well known normalized least mean squares (NLMSs) algorithm as the learning mechanism. However, to correct a nonlinearity, we need a nonlinear mapping/predistorter. This letter proposes a Chebyshev regression-based nonlinear predistorter to correct the nonlinear characteristics of LED by learning a polynomial expansion of the input electrical signal so as to mitigate the LED nonlinearity. Simulations have been carried out to validate the performance of the algorithm against existing adaptive predistortion techniques, such as NLMS-based predistortion and post-distortion techniques, such as Volterra and Hammerstein filters.