Design and Implementation of an Enhanced Load Balancing Algorithm in Fog Computing

Abstract

This study aims to improve load balancing in fog computing networks, which is especially crucial for latency-sensitive IoT applications. An innovative algorithm utilizing Q-learning was designed to optimize load distribution. By incorporating Q-learning agents within fog nodes, the system monitors workloads and resource availability in real-time, enabling adaptive and efficient load balancing decisions. The proposed solution was evaluated using the OMNeT++ simulation framework, yielding significant performance enhancements: a 37% reduction in packet drop rates, a 42% decrease in response times, and a 28% increase in throughput compared to conventional methods.