An intelligent energy minimization algorithm with virtual machine consolidation for sensor-based decision support system

Abstract

Cloud computing has many benefits for businesses because of its distinctive qualities, including scalability, flexibility, on-demand service, and security. A good task scheduler is required to increase the efficiency of a cloud system, which performs numerous jobs simultaneously. On-demand access to resources that have been virtualized is made available as a service without the need for further waiting. For task scheduling issues based on a makespan limitation, energy consumption is decreased, significantly reducing energy cost. Additionally, the complexity of scheduling issues has increased primarily due to the application's lack of a makespan constraint. Unfortunately, reducing the energy cloud services use presents special research issues and difficulties. More precisely, because of the diversity of servers found in cloud centers, it is challenging to choose the best servers for cloud-based decision support systems to reduce energy usage. The presented approach is innovative and could be applied for complex applications while maintaining an average energy consumption for the running resources, which is a big challenge. The current process finds an optimized energy minimization in the cloud with a combination of Virtual Machine consolidation. The outputs were considered in terms of Energy, Virtual Machine Migration, Performance Degradation, Aggregated Ideal Time Factor, and Aggregated Overload Time Fraction. The obtained results in terms of the existing state of art approaches are far better than the competing approaches.

Keywords

• Artificial intelligence
• Cloud computing
• Data analytics
• Intelligent system
• Sensor processing
• Service level agreement virtual machine