Network-aware virtual machine placement using enriched butterfly optimisation algorithm in cloud computing paradigm

Abstract

AbstractThis article presents a virtual machine placement technique aimed at minimizing power usage in heterogeneous cloud data centers. In this study, an innovative model for minimizing the power usage of a datacenter’s network is provided. The Enriched Discrete Butterfly Optimization method (EDBOA) is used as a meta-heuristic method in order to achieve an effective mapping of virtual machines (VMs) onto physical machines (PMs). The Reverse Order Filling Method (ROFM) was developed as a solution repair technique to meet the requirements of the BOA. It is used to manipulate the solutions in order to identify potential candidates for more optimum solutions. Furthermore, we constructed VM’s that had both Left-Right and Top-Down communication capabilities. Additionally, PM’s with limited capacities in terms of CPU, memory, and bandwidth are designed and included for the purpose of testing. The integration of our network power model into the EDBOA algorithms facilitates the calculation of both power modules and network power consumption. A detailed comparative analysis was conducted on our suggested approaches and many other comparable methods. The evaluation findings demonstrate that the offered approaches exhibit strong performance, with the BOA algorithm using the ROFM solution repair surpassing other methods in terms of power usage. The assessment findings also demonstrate the importance of network power usage.