Open Access

ARTICLE

Balancing the load and scheduling the tasks using zebra optimizer in IoT based cloud computing for big-data applications

V. Vijayaraj¹, M. Balamurugan¹, Monisha Oberoi²

1 School of Computer Science of Engineering, Bharathidasan University, Tiruchirappalli, 620023, India
2 Security Services Sales, IBM Innovation Pte Ltd

* Corresponding Authors: V. Vijayaraj (), M. Balamurugan (), Monisha Oberoi ()

Revista Internacional de Métodos Numéricos para Cálculo y Diseño en Ingeniería 2024, 40(2), 1-11. https://doi.org/10.23967/j.rimni.2024.05.009

Received 04 April 2024; Accepted 21 May 2024; Issue published 31 May 2024

Abstract

Task scheduling is one of the major problems with Internet of Things (IoT) cloud computing. The need for cloud storage has skyrocketed due to recent advancements in IoT-based technology. Sophisticated planning approaches are needed to load the IoT services onto cloud resources professionally while meeting application necessities. This is significant because, in order to optimise resource utilisation and reduce waiting times, several procedures must be properly configured on various virtual machines. Because of the diverse nature of IoT, scheduling various IoT application activities in a cloud-based computing architecture can be challenging. Fog cloud computing is projected for the integration of fog besides cloud networks to address these expectations, given the proliferation of IoT sensors and the requirement for fast and dependable information access. Given the complexity of job scheduling, it can be difficult to determine the best course of action, particularly for big data systems. The behaviour of zebras in the wild serves as the primary basis of stimulus for the development of the Zebra Optimisation Algorithm (ZOA), a novel bio-inspired metaheuristic procedure presented in this study. ZOA mimics zebras' feeding habits and their defence mechanisms against predators. Various activities are analysed and processed using an optimised scheduling model based on ZOA to minimise energy expenditures and end-to-end delay. To reduce makespan and increase resource consumption, the technique uses a multi-objective strategy. By using a regional exploratory search strategy, the optimisation algorithm may better utilise data and stays out of local optimisation ruts. The analysis revealed that the suggested ZOA outperformed other well-known algorithms. It was advantageous for big data task scheduling scenarios since it converged more quickly than other techniques. It also produced improvements of 18.43% in several outcomes, including resource utilisation, energy consumption, and make span.

---

Keywords

---