Algorithm for allocating virtual computers in energy-friendly cloud data centers

Abstract

A huge proportion of centres that consume a lot of power, skyrocketing operating costs, and emit a disproportionate amount of carbon dioxide appeared due to the constantly increasing need for processing capacity. This necessitates finding a better approach to handling the energy crisis. This paper proposes a solution to address the energy crisis in cloud data centres, which consume a significant amount of power and emit a large amount of carbon dioxide. A power-conscious simulated engine allocation to physical machines reduces power consumption while maintaining an effective Class of Use. The proposed method is compared with other existing methods such as the Genetic Algorithm (GA), Modified Best Fit Decreasing Algorithm (MBFD), Random Selection (RS), and Minimum Migration Time (MMT) in terms of power usage. The proposed method outperforms the MBFD and GA procedures, showing an 8.55% and 46.81% improvement in power usage, respectively. The MBFD and GA procedures consume 389.5 and 812 kilowatts of energy, while the proposed method reduces this energy consumption. The proposed method is validated using the CloudSim simulator, which demonstrates its success as the fastest and most intelligent procedure compared to other methods. Additionally, the proposed method addresses the burdened host challenge quickly and provides valuable insights for future research by utilizing deep learning approaches.