Multiple Regression Particle Swarm Optimization for Host Overload and Under-Load Detection

Authors

  • Akram Saeed Aqlan Alhammadi
  • Dr. V. Vasanthi

Abstract

Detection of overloaded and under-loaded Host approaches in cloud computing play a vital role. Most of the recent studies use only one resource called CPU to determine the host’s load. In this paper, we propose anaccurate prediction model called Multiple Regression particle swarm optimization (MR-PSO) to detect resource utilization. MR-PSO uses two factors (a) CPU utilization and (b) memory utilization. This model Decreases energy consumption by enhancing the usage of the resources in data centers. The prediction model of the host load based on the Multiple Regression (MR) concept.  Particle swarm optimization (PSO) algorithm is presented to choose the higher and lower threshold borders for host utilization. Simulation by using the CloudSim tool show that MR-PSO decrease the Energy consumption by 7.61% and ESV by 1.5%   lowest than the previous studies when we use the same number of hosts, Virtual machines and tasks.

Downloads

Published

2020-02-19

Issue

Section

Articles