Due to the bandwidth limitation of the ultraviolet-C (UV-C) optical communication system and strong channel attenuation, it is difficult to transmit high-frequency signals. In this paper, the temporal ghost imaging (TGI) algorithm was first applied to the UV-C communication experimentally, and we realized the transmission of a 4 GHz signal through 95.34 MHz system bandwidth. The study indicates that the TGI algorithm can significantly improve the signal-to-noise ratio (SNR) compared with the on–off keying method. Our research provides a new approach for alleviating transmission frequency limitation due to poor SNR and insufficient hardware bandwidth.

In this work, a blue gallium nitride (GaN) micro-light-emitting-diode (micro-LED)-based underwater wireless optical communication (UWOC) system was built, and UWOCs with varied Maalox, chlorophyll, and sea salt concentrations were studied. Data transmission performance of the UWOC and the influence of light attenuation were investigated systematically. Maximum data transmission rates at the distance of 2.3 m were 933, 800, 910, and 790 Mbps for experimental conditions with no impurity, 200.48 mg/m³ Maalox, 12.07 mg/m³ chlorophyll, and 5 kg/m³ sea salt, respectively, much higher than previously reported systems with commercial LEDs. It was found that increasing chlorophyll, Maalox, and sea salt concentrations in water resulted in an increase of light attenuation, which led to the performance degradation of the UWOC. Further analysis suggests two light attenuation mechanisms, e.g., absorption by chlorophyll and scattering by Maalox, are responsible for the decrease of maximum data rates and the increase of bit error rates. Based on the absorption and scattering models, excellent fitting to the experimental attenuation coefficient can be achieved, and light attenuation by absorption and scattering at different wavelengths was also investigated. We believe this work is instructive apply UWOC for practical applications.