An Optimal Host Allocation and Load Distribution Framework Using Maximum Likelihood in Cloud Environment

Abstract

The exponential growth of cloud computing has accelerated the rate of network traffic in data centers. We require effective cloud traffic engineering within each data center in order to conserve storage, energy, bandwidth, and computing power. Therefore, the placement of the virtual machines (VMs) are optimised to take into account the increase in traffic in order to handle the increasing burden in a scalable manner. An optimal host allocation and load distribution framework is proposed called Dynamic Load Balancing based on maximum likelihood (DLB-ML). It announces the concept of bringing the load balancing to achieve the immediate load balancer effect in this dynamic environment. This model inquires the favorable hosts which accomplishes and fulfill the needs of requested tasks. The maximum likelihood estimation is applied to find the probability in terms of computing capability and its performance. The experimental findings demonstrate that our optimal VM allocation method outperforms existing approaches in terms of performance, optimality, throughput and scalability to prove the feasibility and effectiveness of the proposed model. The performance evaluation for dynamic LB model is compared with state-of-art approaches and observe the improvements as well as satisfactory outputs in execution time and resource utilization up to 35.27% and 30.34% respectively in cloud data centers. © 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.