Optimizing Energy Efficient Routing Protocol Performance in Underwater Wireless Sensor Networks With Machine Learning Algorithms

Abstract

Underwater wireless sensor networks (UWSNs) and other communication technology improvements have become increasingly important for monitoring marine environments. These networks predict disasters by analyzing soil properties such as moisture and salinity. The restricted capacity of integrated batteries, along with the challenges associated with their replacement or recharging, has rendered energy efficiency a complex issue in the design of UWSNs. This research suggests a machine learning-based routing protocol that combines the energy-efficient Sea Lion Emperor Penguin Routing Protocol (EESLEPRP) with Gaussian Mixture Clustering (GMCML) to address these problems. The EESLEPRP is used to determine the optimal network path. In this case, the residual energy, delay, and distance of each node is evaluated to determine the optimal path. A comparison shows that the suggested approach yields notable gains, such as a minimal packet loss ratio (PLR) of 2.23%, a 97.76% packet delivery ratio (PDR), and a 90.56% throughput. With an end-to-end latency of 1.38 ms, the model optimizes energy consumption at 97.69%. According to the results, the suggested approach can improve UWSN performance and increase network lifetime.