Underwater wireless sensor networks (UWSNs) have become one of the enabling technologies for the development of future ocean monitoring systems, where autonomous underwater vehicles (AUVs) provide a very attractive way for the data collection of sensor nodes. The data obtained by nodes may have different importance and time sensitivity, which is recorded as the value of information (VoI). The VoI often decays with time. The obtained VoI based on AUV is closely related to the path length of AUV. A longer AUV path lead to more time to collect data, along which less VoI could be obtained finally. Moreover, as sensor nodes are battery-powered and therefore energy-constrained, selecting the nodes with larger residual energy as the AUV path nodes helps to prolong the network lifetime. Therefore, we focus on AUV path planing with the objective of maximizing the lifetime of network while satisfying the constraint on VoI. In this paper, considering the node energy and VoI, we provide an Integer Linear Programming (ILP) model to find an optimal path of AUV. Then, a joint optimized data collection algorithm (JODA) that taking in account the influence of residual energy of nodes and VoI is proposed, which reduces the computational complexity. In our experiments the effectiveness of JODA has been proved by comparing its performance with the theoretical value determined by the ILP model. We also compare the performance of JODA with that of other data collection algorithms, namely TSP, GAAP and EEDA. The simulation results demonstrate that the JODA always outperforms all other algorithms in terms of the total VoI and network lifetime.

水下无线传感器网络 (UWSN) 已成为开发未来海洋监测系统的使能技术之一，其中自主水下航行器 (AUV) 为传感器节点的数据收集提供了一种非常有吸引力的方式。节点获得的数据可能具有不同的重要性和时间敏感性，记录为信息价值（VoI）。VoI 经常随时间衰减。基于AUV获得的VoI与AUV的路径长度密切相关。更长的 AUV 路径导致更多的时间来收集数据，最终可以获得更少的 VoI。此外，由于传感器节点是电池供电的，因此能量受限，选择剩余能量较大的节点作为 AUV 路径节点有助于延长网络寿命。所以，我们专注于 AUV 路径规划，目标是最大化网络的生命周期，同时满足对 VoI 的约束。在本文中，考虑节点能量和 VoI，我们提供了一个整数线性规划 (ILP) 模型来寻找 AUV 的最佳路径。然后，提出了一种考虑节点剩余能量和VoI影响的联合优化数据收集算法（JODA），降低了计算复杂度。在我们的实验中，通过将其性能与 ILP 模型确定的理论值进行比较，证明了 JODA 的有效性。我们还将 JODA 的性能与其他数据收集算法的性能进行了比较，即 TSP、GAAP 和 EEDA。仿真结果表明，JODA 在总 VoI 和网络寿命方面始终优于所有其他算法。