Abstract

This paper evaluates the performance of an underwater visible light communication (UVLC) system employing RGB LEDs, emphasizing the performance gap between ideal deterministic models and realistic stochastic channel conditions. A comprehensive MATLAB-based simulation framework is proposed, integrating Beer-Lambert attenuation with stochastic models for channel fluctuations, pointing errors, and noise. The received power, signal-to-noise ratio (SNR), and bit error rate (BER) are analyzed as key performance metrics under an On-Off Keying (OOK) Intensity Modulation/Direct Detection (IM/DD) scheme across clear, coastal, and turbid water environments. The results demonstrate optimal wavelength selection is heavily dependent on water turbidity, shifting from blue in clear water to green in coastal environments, and red in highly turbid conditions. Furthermore, stochastic effects significantly degrade system performance, reducing the maximum viable communication range by up to 41% compared to deterministic predictions. These findings highlight that mechanical stability and point accuracy are highly influential factors in determining the practical viability of UVLC links beyond theoretical channel limits.