Wavy water-to-air optical camera communication system using rolling shutter image sensor and long short term memory...

Area of Science:

• Optical Communication
• Underwater Technology
• Machine Learning Applications

Background:

• Optical camera communication (OCC) faces challenges in underwater-to-air transmission due to water turbulence.
• Rolling-shutter (RS) technology offers potential for wide field-of-view (FOV) OCC systems.
• Mitigating link outages caused by wavy water is crucial for reliable underwater OCC.

Purpose of the Study:

• To propose and demonstrate a wide FOV water-to-air OCC system using RS technology.
• To evaluate the system's performance under varying water ripple conditions.
• To assess the effectiveness of a Long Short-Term Memory neural network (LSTM-NN) in decoding signals and mitigating turbulence.

Main Methods:

• Experimental demonstration of a novel water-to-air OCC system utilizing RS cameras.
• Implementation of an LSTM-NN for decoding 4-level pulse-amplitude-modulation (PAM4) signals and compensating for water-induced turbulence.
• Evaluation of system performance with different water ripple percentages and camera angular rotations to determine FOV.

Main Results:

• Successful demonstration of water-to-air OCC with RS technology.
• LSTM-NN achieved a pre-forward error correction bit-error-rate (pre-FEC BER) of 3.8 × 10^-3, meeting the required threshold.
• The system supported FOVs of ±70°, ±30°, and ±30° rotations around the z-, y-, and x-axes, respectively, at 6 kbit/s.

Conclusions:

• The proposed RS-based OCC system offers a viable solution for wide FOV underwater-to-air optical transmission.
• LSTM-NN is effective in overcoming water turbulence challenges for reliable OCC.
• The system's wide FOV capabilities open possibilities for diverse underwater communication applications.