Observability Analysis of Underwater Wireless Optical Communication Alignment between AUVs

With the help of underwater wireless optical communication (UWOC) technology, autonomous underwater vehicles (AUVs) can share survey data at high speed during underwater exploration. Establishing a line-of-sight (LOS) link between AUVs and maintaining alignment is necessary for data transmission. We use acoustic navigation to guide two AUVs to maintain a specific relative position and direction for beam alignment. The relative motion and acoustic observation between two AUVs are modeled in this research. The local weak observability matrix is presented to analyze the observability in the alignment process. To improve observability, the heuristic method and reinforcement learning method are utilized for motion planning. The reinforcement method outperforms the heuristic method in simulation experiments, which improves observability and reduces the use of acoustic resources.