Signal Decision Employing Density-Based Spatial Clustering of Machine Learning in PAM-4 VLC System

摘要

As light emitting diode (LED) based visible light communication (VLC) is getting increasingly widely used, amplitude jitter is still a common phenomenon in pulse amplitude modulation (PAM) VLC system, which deteriorates the system performance to a large extent. In this paper, we propose a novel signal decision method employing density-based spatial clustering of applications with noise (DBSCAN) of machine learning to distinguish different signal levels with jitter. Not only do we experimentally demonstrate that the Q factor of a PAM-4 VLC system employing DBSCAN is improved by up to 3.9dB, but also investigate the influence of jitter with different levels on PAM-4 system. As far as we know, this is the first time that DBSCAN has been successfully employed in PAM-4 VLC system.