Intelligently Routing Underwater Wireless Sensor Networks: A Survey

Abstract

Current marine scientific research faces a big challenge in accurately and timely collecting marine data due to the highly dynamic underwater environment. Underwater Wireless Sensor Networking (UWSN) systems with intelligent routing approaches are a way to rise to this challenge as the rapid development of Artificial Intelligence (AI) technologies. Unlike traditional routing approaches driven by mathematical-models, the ones underwater are intelligent and data driven, which better adapt to the dynamic underwater environment and are able to optimize various performance indicators. This paper reviews the state of arts of applying AI technologies in routing UWSN systems. It divides the routing approaches for UWSN systems into non-cross-layer and cross-layer based. Cross-layer based approaches can be further grouped into traditional cross-layer and intelligent cross-layer approaches, as AI technologies can help data sharing among layers in UWSN systems. This paper focuses on the intelligent cross-layer routing approaches and discusses them according to the type of machine learning algorithms applied. Challenges and some open issues in intelligently routing UWSN systems are also provided. To the best of our knowledge, this work is the first effort on summarizing the challenges, algorithms, and issues in routing UWSN systems intelligently.