Load Balancing in Cloud Environment using Stackelberg's Approach

Abstract

Cloud Computing, one of the latest standard of large scale and parallel computing, attracts more users requiring utility computing with better and fast service. Load balancing is one of the key parameters that determine a cloud data center's performance. One of the main problems with respect to load balancing is that all the physical hosts in a data center are not efficiently used resulting in an imbalance and suboptimal performance. This paper has focused on how the physical hosts for deploying requested tasks is selected based on the requirement of the request by using Stackelberg's approach. Most of the works that has been done previously in this area utilize a series of algorithms that selects the optimal host in the data center based on the intelligence that is confined to the algorithm alone and doesn't have a holistic approach that considers the unutilized resources amongst all hosts within a data center. The proposed model in this work makes a decision of which physical host must be allocated to a requestbased on the requirements of the incoming task, current load on the data center hosts, available or unused resources in the data center. The model uses First-in-First-out allocation strategy for task assignments. Simulation results compared with the existing works show that the proposed approach has decreased the failure number of task deployment events obviously, and reduced the makespan.