Dynamic Load Balancing in Cloud-IoT Based on Neural Network Classification and GGA Optimization

The Cloud IoT paradigm, designed to combine Cloud Computing (CC) and the Internet of Things (IoT) benefits, is increasingly utilized for extensive services and addressing users' connectivity, data processing, and storage needs. However, achieving load balancing in Cloud-IoT setups remains challenging due to dynamic and diverse connected objects and Cloud resources. This paper presents a novel approach that employs neural networks for task classification from connected objects, coupled with a task scheduler merging genetic and grasshopper optimization algorithms. This aims to dynamically balance loads in the Cloud-IoT environment. The proposed method is compared with recent alternatives, assessing metrics like makespan, throughput, resource use, energy consumption, and cost. Experimental results compellingly demonstrate the superiority of our approach across these factors.