Efficient Firefly Algorithm based Energy Saving Technique in Cloud Data Center
Abstract
The increasing demand for energy by cloud data centers has led to the generation of large amounts of energy, which results in high operating costs and emission of . In addition, it becomes necessary for the cloud computing providers to provide high-quality services to their users and hence need to manage power shortage. Reducing energy consumption by large data center becomes a great challenge for researchers as well as scientists. To solve this problem, VM migration methods are adopted for server consolidation to minimize energy consumption. In this research, the work is carried out into three phases: In the first phase, the energy utilization of each PM is analyzed and then arranged them in decreasing order according to their energy consumption using Modified Best Fit Decreasing (MBFD) approach. In second phase, Firefly (FF) algorithm is applied to deal with the new user requests. In the last phase, the formerly deployed and assigned VM’s are ratified in terms of normal loaded, overloaded or under-loaded by adopting artificial neural network (ANN) approach. Using the proposed technique, the host /machine that has minimum possibility of over-utilization and needs least migration is selected as the best host for migrating VM. The performance of the proposed work has been exhibited in MATLAB simulator and validated the results in terms of number of hosts, number of migrations and energy consumptions. The results show that energy up to 43.9 % is saved, thereby, making the cloud data Centre more energy aware.