Design of Heterogeneous Multi-AUV IoUT: Architecture, Key Technologies and Applications

Internet of underwater things (IoUT) has been widely used in developing marine resources, in which autonomous underwater vehicles (AUVs) are often employed as task execution units. In this article, we propose the concept of heterogeneous multi-AUV IoUT, which elegantly integrates centralized and distributed AUVs to meet the needs of complex underwater tasks. Specifically, we introduce the air-shore-sea integrated network with a layered architecture, which can support the communication and transmission of data and instructions in the heterogeneous multi-AUV IoUT. Under this architecture, we develop synthesis heuristic dynamic routing to support the underwater cross-layer information transmission of centralized AUVs and thus improve wireless communication quality. We adopt multi-agent deep reinforcement learning (MADRL) technology to realize the automatic control of distributed AUVs in multiple underwater tasks. Furthermore, we discuss the multi-angle approach to IoUT security management. Finally, the open questions and potential future research trends are investigated.