Asynchronous Localization with Stratification Effect for Underwater Target: A Reinforcement Learning-based Approach

In this paper, we are concerned with the localization of underwater target under the asynchronous clock and stratification effect. A network architecture is established that comprises of surface buoys, sensor nodes and the target. Sensor nodes act as anchor nodes and communicate with target. With the collected localization messages, the relationship of time differences and propagation delay is established. Then the reinforcement learning-based approach is designed to solve the localization optimization problem. The value iteration process is given to determine the optimal policy. Finally, simulation results are presented to show the effectiveness of the proposed method.