Enhanced Artificial Bee Colony Algorithm for Load Balancing in Cloud Computing Environment

Abstract

IT companies are moving their business from traditional computing to cloud computing to leverage the benefits like reduced infrastructure cost, business agility, etc. This change has led to an increase in the number of cloud-based service users worldwide from 2.4 billion in 2013 to 3.6 billion in 2018. By the end of 2020, this number will be double of 2013 users. The increased number of users, in turn, has increased the number of accesses done to web servers. Increased number of dynamic requests can create problems like reduced system performance, system break down, inefficient utilization of resources, etc.; All these problems can be avoided by distributing the workload among all resources using the load balancer. In this paper, a bio-inspired load balancing algorithm named Artificial Bee Colony is enhanced and compared with the basic ABC algorithm. The enhancement is done by merging ABC with Capacity Based Load Balancing algorithm. The experiments are carried out in the CloudSim simulator as well as real cloud environment using Amazon Web Services (AWS) platform. The significance of the results is tested by using statistical tests. The performance of the proposed algorithm is analyzed for make span, throughput, response time and the average waiting time.