SBEIN: Toward Sustainable Battery-Efficient IoT-Enabled Federated Learning Networks

The convergence of the Internet of Things (IoT) and Federated Learning (FL), driven by advancements in 5G, presents significant opportunities for Consumer IoT (CIoT) applications. However, despite this progress, CIoT-enabled FL systems face critical challenges in maintaining energy efficiency and optimal system performance, particularly in selecting consumer devices that balance power consumption and processing capabilities. The SBEIN framework dynamically optimizes device selection and power management using Adaptive Particle Swarm Optimization (PSO) and a modified Genetic Algorithm (GA). Experimental results show that SBEIN reduces energy consumption by 20% and extends battery life by 30%, enhancing sustainability and user experience in CIoT-driven FL environments.