سازوكار تجميع ماشين¬هاي مجازي براي بهبود تعادل بار و بهره¬وري انرژي در مراكز داده ابري ناهمگن

The Internet of Things has revolutionized various fields, including healthcare, transportation, and urban management, by connecting objects to the global network. IoT devices generate vast amounts of data, surpassing their internal processing and storage capacities. To effectively manage and extract value from this data, cloud computing emerges as a powerful solution. However, the rapid growth of data centers to support cloud computing raises concerns about increased energy consumption and environmental impacts. Virtual machine consolidation is proposed as an efficient approach to reduce energy consumption and optimize resource utilization. One of the key challenges in virtual machine consolidation is improving energy and resource efficiency without compromising service quality and violating service level agreements. This research proposes a comprehensive approach to simultaneously address these challenges. The approach is executed in three stages: first, the future consumption of physical machines is estimated using a predictive algorithm, and their load status is identified. Next, suitable virtual machines for migration are selected using a resource-aware algorithm, with the selection based on resource skewness analysis and weighting. Finally, a hybrid algorithm (heuristic and energy-aware) is used to choose an appropriate physical machine to host the virtual machines, also considering load balancing across different resources. The eva‎luation of the results shows that the proposed approach effectively achieves its goals, including reducing energy consumption without degrading service quality and violating service level agreements. Compared to a similar approach, the proposed approach shows improvements of 24% in energy consumption, 3.4% in number of virtual machine migrations, 35.5% in SLAV, 51% in ESLAV, and 52.7% in ESM

6