A BIO-INSPIRED APPROACH TO TASK SCHEDULING IN FEDERATED CLOUDS USING WALRUS OPTIMIZATION ALGORITHM
No Thumbnail Available
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Description
Cloud computing has witnessed an exponential rise in popularity, leading to a substantial increase
in cloud users. This surge in demand for cloud services has introduced significant challenges in
delivering high-quality services and optimizing resource allocation. As the demand for cloud
services expands, effective task scheduling becomes paramount for improving system
performance. Current task scheduling approaches are limited in that they cannot guarantee finding
the globally optimal solution for optimization problems; often, they settle for locally optimal or
suboptimal solutions, leading to underutilization of resources, increased expenditure, and customer
dissatisfaction. This research investigates the potential of the Walrus Optimization Algorithm
(WaOA) for task scheduling in federated clouds. The algorithm's performance was compared with
existing approaches using standard metrics such as makespan, execution time, and throughput
across various workload scenarios. The study utilizes Java and CloudSim for implementation and
evaluation. Results demonstrate WaOA's efficiency in enhancing task scheduling within federated
clouds, achieving the shortest makespan, highest throughput, and lowest execution time among the
algorithms tested. Its ability to adapt to dynamic environments and optimize resource utilization
consistently proved valuable across diverse scenarios. As the number of data centers increased,
WaOA consistently performed well, indicating its potential for handling larger workloads and
improving resource utilization efficiency. Overall, the study concludes that WaOA is a promising
solution for enhancing task scheduling efficiency in federated clouds, highlighting the significance
of algorithm selection and data center configuration in cloud computing environment.
Keywords
QA Mathematics, QA75 Electronic computers. Computer science