Dynamic Layered Intelligent Underwater Acoustic Data Collection Scheme Based on Value of Information and Residual Energy

The technology of underwater acoustic sensor networks (UASNs) for collecting underwater acoustic data has been advancing rapidly. To address the challenges of short network lifetime, imbalanced energy consumption, and long end-to-end delay, researchers have been developing innovative methods of path planning for data collection and protocols for data transmission. In this article, we propose an underwater acoustic data collection scheme that simultaneously considers both the value of information (VoI) and the average residual energy of the network. When the VoI is high and the average residual energy is adequate, a multihop transmission mode is utilized for data collection, while in other cases, autonomous underwater vehicles (AUVs) are deployed to collect the data. This dual strategy ensures timely data acquisition while extending the network’s lifespan. In addition, we develop a dynamic layered data collection approach for AUVs based on cluster density and depth. The Q-learning algorithm is employed to adaptively optimize AUVs path planning across layers, thereby reducing the traveling time and improving the VoI. Simulation results demonstrate that our proposed method effectively prolongs the network’s lifetime and enhances data collection efficiency.