Hybrid Markov chain-based dynamic scheduling to improve load balancing performance in fog-cloud environment

Abstract

Fog computing is a distributed computing paradigm that has become essential for driving Internet of Things (IoT) applications due to its ability to meet the low latency requirements of increasing IoT applications. However, fog servers can become overburdened as many IoT applications need to run on these resources, potentially leading to decreased responsiveness. Additionally, the need to handle real-world challenges such as load instability, makespan, and underutilization of virtual machine (VM) devices has driven an exponential increase in demand for effective task scheduling in IoT-based fog and cloud computing environments. Therefore, scheduling IoT applications in heterogeneous fog computing systems effectively and flexibly is crucial. The limited processing resources of fog servers make the application of ideal but computationally costly procedures more challenging. To address these difficulties, we propose using an Arithmetic Optimization Algorithm (AOA) for task scheduling and a Markov chain to forecast the load of VMs as fog and cloud layer resources. This approach aims to establish an environmentally load-balanced framework that reduces energy usage and delay. The simulation results indicate that the proposed method can improve the average makespan, delay, and Performance Improvement Rate (PIR) by 8.29 %, 11.72 %, and 4.66 %, respectively, compared to the crow, firefly, and grey wolf algorithms (GWA).