부산대학교 도서관

Nonlinear optoacoustics enable effective communication across the air-water interface. However, the requirement of a high-power laser and the vapor cloud buildup can limit the power efficiency and data rate. Thus, a proper modulation and encoding scheme is necessary. This article tackles this issue by presenting an optical focusing-based adaptive modulation (OFAM) technique that can dynamically control the underwater acoustic source (plasma) and acoustic pressure. Specifically, the article describes two variants of OFAM for a single laser transmitter with stationary focusing (OFAM-1D) and dynamic focusing (OFAM-3D). The data rate of OFAM-1D and OFAM-3D is approximately 6 and 4.4 times higher than peak detection based on-off keying (PDOOK). Furthermore, both techniques are 137% more power efficient than PDOOK. Studying the bit error rate (BER) in the presence of ambient underwater noises for different node positions has indicated that OFAM can achieve low BER even at a 300-m depth for 50- and 60-mJ laser pulse energy. Moreover, machine learning (ML) techniques have been leveraged in the demodulation process for increased robustness. Specifically, the random forest (RF) model could yield up to 94.75% demodulation accuracy. Our results indicate that OFAM can lead to a new paradigm of air to underwater wireless communication.