Virtual Machine Placement Using Hypercube Ant Colony Optimization Framework

Abstract

This paper presents a novel application of hypercube framework of ant colony optimization to the virtual machine (VM) placement problem, with the objective of minimizing the power consumption and resource wastage. The aim of the VM placement is to develop an optimal placement strategy to allocate the VM's to physical servers such that the usage of physical resource is minimized. The hypercube framework of ACO differs from its usual ACO implementation in the hyperspace for the pheromone values. In other words, in the hypercube framework, the updating of pheromone values is constrained to lie between 0 to1. This constrained pheromone updating has an advantage of automatically handling the scaling of objective function values and further leads to robust version pf ACO procedure. Here, we experimentally investigate the influence of hypercube framework of ACO for virtual machine placement using three ACO variants, namely, Ant system, Max-Min Ant System, and Ant colony system.

Meta-heuristic Techniques for Placement of Virtual Machines in Distributed Cloud Centers

“…In [22] authors developed system to minimize power consumption and resource wastage in virtual machine placement using ant colony optimization with hypercube framework. In [23] generalized online optimization methodology was used to place as many virtual machines as possible in both nondistributed and distributed cloud data centers to increase the cloud service provider's revenue.…”