Energy-Efficient Task Scheduling Using a Novel Metaheuristics Approach in Cloud Data Centers

Cloud computing operates under a centralized computing framework and employs a distributed computing paradigm that revolutionizes delivering computing services. With this transformative technology, users can access virtualized and elastic resources, including computation, platforms, and applications software, which can be provisioned and released instantaneously according to demand. The cloud infrastructure, comprising numerous data centers, requires substantial power consumption to efficiently deliver services to users on demand. The increased power consumption of these resources translates to higher costs for service provision. Current task scheduling methods predominantly focus on objectives like load balancing or makespan, but most of these approaches neglect to optimize power consumption for heterogeneous cloud platforms. The proposed EETSJMSPSO is a new novel metaheuristic approach known as Energy Efficient Task Scheduling using JAYA and Multi-Swarm Particle Swarm Optimization. The primary goal of EETSJMSPSO is to address the persistent challenge of high-power consumption when virtual machines are not subject to migration. The proposed algorithm allocates tasks intending to optimize energy consumption and makespan by implementing load balancing. It also aims to outperform existing schemes dynamically by optimizing other quality of service (QoS) parameters.