A Novel Maritime Edge Computing System for Internet of Vessels

In this paper, a novel maritime edge computing system is proposed and demonstrated. In the proposed system, a cloud-edge-end architecture is deployed, and computing functions are included at both the “edge” and “end” sides. Therefore, the computing capabilities of the edge computing system are enhanced by increasing the number of computing nodes and assigning tasks to different computing locations. The system adopts Unmanned Surface Vehicles (USVs) and buoys as the edge computing carriers at the “end” side, and they are equipped with video acquisition cameras, hydrological information sensors, supporting edge computing units and communication units to enable information collection, computation and task uploading. At the same time, the “edge” refers to the shore and islands near the USV, and is equipped with more powerful edge computing and communication units to realize the collection and computation of the broadband data that cannot be computed at the “end” side, as well as uploading computing results to the cloud. Finally, the “cloud” is the client which centrally displays the real-time video information, hydrological information and computing results from both the “edge” and “end” sides to provide the support for decision makers. In addition, communication units in the proposed system can establish a flexible communication network to ensure that the nodes in the edge computing system are always online, which provides a guarantee for information transmission. Finally, we trained the edge computing unit of the system and implemented field experiments on multi-object identification under various scenarios to verify the feasibility of the system. The system can provide new solutions for maritime rescue and exploration, and has an important significance on ocean digitization and intelligence.