Fair Distribution Routing Algorithm for Underwater Wireless Communication using Machine Learning

The underwater wireless sensor network (UWSN) is a wireless system used in aquatic environments. Unlike terrestrial wireless networks that rely on radio frequency (RF), UWSNs use alternative communication methods like acoustic channels, due to RF's incompatibility with underwater conditions. The dynamic nature of UWSNs, where sensors move with water activity, leads to rapidly changing network topologies. A significant challenge is efficiently collecting and routing data from distributed sensors to a central sink node. Traditional terrestrial routing protocols are not directly applicable to UWSNs. This paper introduces a fair distribution routing protocol (FDRP) for UWSNs, which addresses power limitations, latency, and void area issues using reinforcement learning (Q-learning). The Fair Distribution Routing Protocol (FDRP) outperforms the QELAR protocol’s performance. Simulations showed that FDRP reduced latency by 18%, allowing for faster data transmission, while also improving energy efficiency by 16%, meaning it consumes less power while maintaining or surpassing the performance of QELAR. These improvements highlight FDRP's effectiveness in optimizing signal propagation and data routing, leading to faster communication and reduced energy consumption.