Minimization of Energy Consumption for VM Placement using Normalized Three Dimensional Resources

Abstract

Cloud computing is used for resource management in a virtualized environment. The Virtual Machine (VM) placement is important for better utilization of resources. That leads to reduce energy consumption. Inefficient allocations are observed for traditional 1D bin-packing methods where multidimensional VM resources are available. This study discussed an enhanced VM placement strategy based on 2D and 3D bin packing, which incorporates CPU and RAM as key resource dimensions. An improved method for optimizing VM allocation decisions is developed by normalizing these parameters. The experiments are conducted on CloudSim, which evaluates the proposed approach against a traditional 1D first-fit algorithm in terms of key performance metrics such as energy consumption, VM migrations, and execution time. The results highlight the advantages of multi-dimensional bin packing and demonstrate superior efficiency and resource utilization. It has been observed that energy consumption for 2D bin packing approach reduces by 2.87% as compared to 1D bin packing. Further, it reduced by 5.52% from 2D to 3D bin packing. This shows a significant overall reduction. This study shows that there can be further improvements in high-dimensional VM placement strategies.