Nonlinear adaptive equalization with data selection in VLC systems

Abstract

This work proposes the use of data-selective equalizers for visible light communication (VLC). In such systems, one of the main sources of nonlinearity is the light-emitting diode. Nonlinear techniques such as Volterra series are employed to cope with this effect. However, the large number of parameters underlying to the model yielded by such series grows considerably the computational complexity. This work proposes the use of data-selective adaptive schemes in Volterra series. By working only with input data that brings novelty to the system, thus avoiding unnecessary parameter updates, one can reduce drastically the high computational burden associated with the use of Volterra series, therefore, providing an efficient nonlinear equalizer. In order to reduce even more the computational burden, this work also proposes the combination of different data-selective algorithms. Thus, by exploring the structure of the Volterra series, one can employ a filter for each of its kernels. Each of these filters update more elements as the order of the kernel increases. Therefore, by assigning larger error thresholds to these filters, it is possible to decrease their update rates, thus reducing the overall computational complexity. Also, it is proposed the use of data-selective semi-blind algorithms to decrease the amount of data used to train the adaptive filters, while reducing the computational complexity. A computational platform for simulation of VLC systems is also one of the proposals of this work. The key components of this system are presented, whose models are based on the description of their physical properties performed in this work.