In this paper, we consider a underwater wireless sensor network, which is composed of an autonomous underwater vehicle (AUV) and the underwater data collection stations. Constrained by limited resources, at each decision epoch the AUV selects one underwater data collection station to update the data and then uploads it to the surface base station. Different from other underwater communication models, we consider the influence of underwater environment on the AUV and develop an AUV model with random lifetime, meanwhile, we further propose that the information arrives randomly with different importance levels. The system model is formulated into a discounted Markov decision process with infinite horizon. Taking the age of information and the importance of information as the performance metrics, the optimal scheduling strategy of the AUV is developed, and the threshold structure characteristics of the optimal strategy are proved. Finally, the numerical examples are given to illustrate the theoretical results, and the impacts of system parameters on the optimal strategy and the value function are also investigated.

在本文中，我们考虑了一个水下无线传感器网络，它由自主水下航行器（AUV）和水下数据采集站组成。受限于资源，AUV在每个决策时期选择一个水下数据采集站更新数据，然后上传到水面基站。与其他水下通信模型不同，我们考虑水下环境对 AUV 的影响，并开发了具有随机寿命的 AUV 模型，同时，我们进一步提出信息以不同的重要性级别随机到达。系统模型被制定为具有无限视野的贴现马尔可夫决策过程。以信息时代和信息重要性作为性能指标，制定了AUV的最优调度策略，并证明了最优策略的阈值结构特征。最后通过数值算例说明了理论结果，并研究了系统参数对最优策略和价值函数的影响。