Virtual Machine Placement in Cloud Environments Using a Hybrid Cuckoo Search and Bat Algorithm

Abstract

The growing popularity of on-demand pay-as-you-go subscription models for online cloud computing requires increasing amounts of resources to ensure adequate quality of services. However, to satisfy the strong demand for these services, cloud infrastructure providers continue to scale up their data centers. This scaling often lacks an optimal resource management approach, thus leading to inefficiencies, excessive energy consumption, and higher costs. This creates challenges in the virtual machine placement (VMP) process focusing on identifying efficient ways for assigning virtual machines to physical hardware. This paper introduces a hybrid cuckoo search bat algorithm (HCS-BA) to solve VMP in heterogeneous cloud environments. The suitability of the cuckoo search algorithm for global searches is combined with the local refining capacity of the bat algorithm, therefore optimizing both energy consumption and resource utilization. The results of simulations carried out in Matlab and CloudSim for scalability testing demonstrate that HCS-BA outperforms both individual algorithms. It reduces energy consumption and improves resource utilization.