Optimizing Cloud Computing Performance With an Enhanced Dynamic Load Balancing Algorithm for Superior Task Allocation

Raiymbek Zhanuzak, Mohammed Alaa Ala’anzy, Мohamed Othman, Abdulmohsen Algarni

IEEE Access, 2024

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.1109/access.2024.3508793
Data confidence 85 high
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Cloud computing, particularly within the Infrastructure as a Service (IaaS) model, faces significant challenges in workload distribution due to limited resource availability and virtual machines (VMs). Efficient task allocation and load balancing are crucial to avoiding overloading or under-loading scenarios that can lead to execution delays or machine failures. This paper presents an Enhanced Dynamic Load Balancing (EDLB) algorithm designed to optimise task scheduling and resource allocation in cloud environments. Unlike benchmark algorithms that rely on static VM selection or post-hoc relocation of cloudlets, the EDLB algorithm dynamically identifies optimal cloudlet placement in real-time. Our approach proactively allocates cloudlets to VMs based on current system states and Service Level Agreement (SLA) deadlines, thereby preemptively addressing potential SLA violations. Additionally, if a VM cannot meet the deadline of the cloudlet, the algorithm redirects the cloudlet to a secondary data centre and reconfigures CPU resources among VMs to ensure optimal allocation. Evaluations using CloudSim simulations demonstrate that the EDLB algorithm achieves substantial average improvements over benchmark algorithm and the-state-of-the-art algorithm, including a 59.46% reduction in total makespan, a 12.70% reduction in average makespan, a 22.46% reduction in execution time, and a 3.10% increase in resource utilisation. Furthermore, the EDLB algorithm enhances load balancing by 46.46%. These results highlight the effectiveness of the EDLB algorithm in addressing critical load balancing issues and surpassing existing methods. This research contributes to the field by introducing a novel approach that significantly improves performance metrics and operational efficiency in cloud computing environments.