Abstrato

Localization Through Optical Wireless Communication in Underwater by Using Machine Learning Algorithms

Fnu Ziauddin*

With respect to ocean exploration, underwater robots, and environmental supervision, accurate localization in an underwater setting is perhaps among the principal hurdles that need to be overcome. For underwater localization, conventional acoustic-based methods usually suffer from high latency, short communications’ ranges, and vulnerability to multipath fading. The paper proposes a new approach to precise and successful positioning under water that exceeds specified constraints by pairing state-of-the-art artificial intelligence procedures using Optical Wireless Communication (OWG). In this study, we propose the establishment of a hybrid architecture that merges the advantages of LSTM networks, RNNs, SVMs, and CNNs. This integrated system is meant to process and examine efficiently the temporal evolution of underwater optical signals. The paper has proposed a dual-mode communication strategy that uses optical techniques like signals for transmission within the visual range and audio technologies for broadcasts out of line of sight. The hybrid optical-acoustic system acts as one of the key improvements towards better data transfer, position tracking, and underwater communication by blending the advantages of two different technologies. In this regard, reliable communication along the entire optical connection path length is highlighted, especially in relation to the continually changing target’s dynamics. This machine learning with OWCenabled localization method performs significantly better as compared to conventional acoustic-based approaches in terms of reduced latency, enhanced communication ranges, and improved positioning accuracy. In addition to this, the present work may be the first to reveal a considerable step forward in precise and fast subsea positioning, as well as the prospect of expanding the technologies for underwater inspection and interaction.