Enhanced Cloud Load Balancing With MPSOA- LB: A Multi- Objective PSO Approach for Dynamic Task Allocation and Performance Optimization

Abstract

Abstract—Cloud computing is a computing environment that involves the process of accessing of services, which
includes the storage environments, various applications and servers through the Internet. Cloud is a large pool of
information, software packages, shared resources, storage and enormous applications based on user demands at any point
of time. In other words, cloud computing refers to system-oriented software, physical hardware devices, and day-to-day
applications delivered to the users through the medium of the internet as services. Cloud resources can be accessed in
diverse ways for multiple purposes, making scheduling crucial for providing optimal services to users. With data and
resource availability increasing constantly, the need for efficient scheduling algorithms becomes paramount. Effective
load-balancing techniques can significantly enhance system performance while reducing costs and energy consumption.
Various heuristic algorithms have been proposed to tackle these challenges, with intelligent approaches being widely
adopted. This paper portrays the detailed description of the MPSOA-LB scheme propounded for attaining substantial load
balancing in a cloud computing setting. The planned system model explores various factors that contribute to the
development of a fitness function, which helps evaluate the over-utilization and under-utilization in the MPSOA-LB. This
algorithm focuses on efficient load distribution among virtual machines and hosts within cloud environments. The paper
also discusses the simulation setup, and the results obtained from implementing the MPSOA-LB under varying conditions,
including the quantity of tasks, instruction finishing lengths, and increasing the quantity of virtual machines.