Machine Learning Aided Signal Detection in Underwater Wireless Optical Communication for IoUT applications

This study explores the effectiveness of the Machine Learning (ML) algorithms in addressing channel impairments in underwater Wireless Optical Communication (UWOC) systems employing On-Off Keying (OOK) modulation. The simulation takes into account underwater challenges such as signal attenuation, scattering, and absorption by using the Gamma-Gamma distribution to model fading and scintillation effects. The ML algorithm's performance is assessed by comparing its bit error rate (BER) against the signal-to-noise ratio (SNR) in comparison to the ideal scenario with perfect channel state information (CSI). The simulation covers various underwater scenarios, including ocean water, harbor water, clear water, coastal water, and oligotrophic water, showcasing the algorithm's adaptability in diverse environmental conditions. The results indicate that the SVM algorithm closely approaches the BER performance achieved with CSI, demonstrating its potential to improve communication reliability in UWOC systems under realistic channel conditions. This study provides valuable insights into the application of machine learning for signal detection in UWOC, offering prospects for enhanced underwater communication performance.