LED可见光通信系统中应用基于密度的聚类算法的信号判诀

Signal Decision Employing Density-Based Clustering in LED-Based Visible Light Communication System

在可见光通信系统中,无载波幅度相位(carrierless amplitude/phase, CAP)调制是国际主流的调制方式之一。但是受噪声的影响,幅度的不稳定抖动会大幅降低系统性能。针对这个问题,本文提出了一种结合机器学习的基于空间密度的群以噪声发现聚类算法(density-based spatial clustering of applications with noise, DBSCAN)的信号判决方法,来区分CAP-4系统中的4个带抖动的不同电平。通过应用该方法进行实验,可见光通信CAP-4系统的性能得到了高达4.77 dB的提升。同时,本文还研究了不同幅度的抖动对系统的不同影响。这是DBSCAN算法在LED可见光通信CAP-4系统中的创新应用。

Carrierless amplitude/phase(CAP) modulation is one of the most internationally used modulation format in visible light communication(VLC) system. However, due to the influence of noise, the unstable jitter of amplitude will deteriorate the system performance to a large extent. To solve this problem, this paper proposes a signal decision method employing density-based spatial clustering of applications with noise(DBSCAN) of machine learning to distinguish four different signal levels with amplitude jitter in CAP-4 system. The Q factor of CAP-4 VLC system is improved by up to 4.77 dB and the influence of jitter of different amplitude on CAP-4 system is also investigated. As far as we know, this is the innovative way that DBSCAN has been successfully employed in CAP-4 VLC system.