A METHOD BASED ON MODIFIED SIMULATED ANNEALING AND OPTIMAL VECTOR FOR LOAD BALANCING IN FOG COMPUTING

Abstract

Fog computing is a distributed and hierarchical framework designed to deliver services such as computing, storage, and network resources. It reduces latency and communication frequency between users and edge nodes, offering a solution that minimizes delays and traffic congestion, while also lowering power consumption and bandwidth usage. This helps alleviate the burden on cloud data centers. While cloud computing remains a practical solution for sustainable development, particularly in the IoT sector, it still faces several unresolved challenges. One of the key issues in fog computing is load balancing, which plays a crucial role in system management. A major challenge is selecting the most suitable source for each task. Load balancing in cloud computing involves distributing workloads across multiple nodes to ensure high resource utilization, equitable resource allocation, and user satisfaction. In this work, simulated annealing and optimal vectors are used to achieve load balancing in fog computing. The simulated annealing algorithm helps identify the best values for the request allocation vector, ensuring tasks are efficiently assigned to servers. The method was implemented in MATLAB, and the results demonstrate notable improvements in average response time, total execution time, and load balancing performance.