SVM aided LEDs selection for generalized spatial modulation of indoor VLC systems

Abstract In order to reduce the complexity of the light-emitting diodes (LEDs) selection procedure in generalized spatial modulation (GSM) assisted indoor visible light communication (VLC) system, a support vector machine (SVM) aided low complexity and high efficiency machine learning LEDs selection algorithm is proposed for the considered GSM–VLC system. By modeling the LEDs selection problem in indoor GSM–VLC system as a multi-classification task, an optimization problem is constructed by utilizing kernel SVM. After the optimal parameters are obtained from the training stage, an LEDs selection procedure can be accomplished efficiently by SVM aided learning system for any given user’s channel state information. Simulation results and complexity analysis show that, compared with traditional LEDs selection algorithms, the proposed SVM aided LED selection algorithm can achieve an ideal bit error ratio (BER) performance while having considerable lower complexity for the considered GSM–VLC system.