Machine-Learning-Aided Mission-Critical Internet of Underwater Things
Xiangwang Hou, Jingjing Wang, Zhengru Fang, Xin Zhang, Shenghui Song, Xudong Zhang, Yong Ren
Abstract
With people paying more attention to marine resources, the Internet of Things (IoT) has been extended to underwater, promoting the development of the Internet of Underwater Things (IoUT). Various compelling IoUT applications bring a new age to maritime activities. However, some mis-sion-critical maritime activities, including ocean earthquake forecasting, underwater navigation, and so on, pose a substantial challenge on existing IoUT architectures and relevant techniques. Therefore, in this article, to empower these implacable maritime activities, we conceive the concept of mission-critical IoUT and highlight its key features and challenges. Furthermore, to satisfy the stringent requirements of mission-critical IoUT, we propose a future maritime network architecture and machine-learning-aided key techniques in terms of information sensing, transmission, and processing. Moreover, we present our recent research on reliable and low-latency underwater information transmission. Finally, we provide the open issues and potential research trends for future mission-critical IoUT.