FULL PDF
Abstract
Federated Edge Intelligence (FEI) presents a transformative approach for enabling real-time analytics and decision-making within the Internet of Things (IoT) ecosystem, particularly under resource constraints. By decentralizing machine learning and pushing computation toward the edge, FEI minimizes latency, preserves privacy, and reduces communication overhead. This paper explores key advancements in integrating federated learning into resource-constrained IoT frameworks, identifies existing challenges such as energy efficiency, model accuracy, and heterogeneity, and suggests architectural strategies to overcome these barriers. The study concludes with prospective directions, emphasizing edge-cloud synergy, energy-aware learning, and adaptive model deployment for robust and scalable IoT solutions.
View more >>