A Power Control-Aided Q-Learning-Based Routing Protocol for Optical-Acoustic Hybrid Underwater Sensor Networks

Underwater optical-acoustic hybrid networking can complement the advantages of both underwater wireless optical and acoustic communications, which holds great development prospects. This paper designs a novel and feasible underwater optical-acoustic hybrid sensor network topology, which fully considers the differences between underwater sound and light communication systems. Following this network topology, an adaptive and green power control aided routing protocol is proposed, mainly including two stages: the power control and the packet forwarding. These two procedures employ potential game theory and
${Q}$
-learning technique respectively, making each node adjust its transmitted power and forward packets in an adaptive and distributed manner with energy efficiency, packet advancement, and network connectivity as the optimization goals. According to the simulation results, the proposed power control and packet forwarding algorithms are of reasonable complexity and convergence speed, and adding the power control before the packet forwarding can achieve better routing performance. Compared with relevant routing protocols, the designed protocol has significant advantages in energy efficiency, end-to-end delay, and network lifetime.