Cuckoo Search inspired VM migration and allocation based on Neural Architecture

  • Kanav Sadawatia, Satish Saini
Keywords: VM allocation, VM migration, Cuckoo Search (CS), Artificial Neural Network (ANN).

Abstract

In the current scenario, with the rising popularity of cloud computing, the network virtualization is gaining attention thorough out the world. The network virtualization functions to balance the work load of servers by distributing the task among various virtual machines without compromising Service Level Agreement (SLA) conditions. However, this idea has an adjoining challenge of energy consumption. In the present paper, authors had proposed a Cuckoo Search (CS) optimized Virtual Machine (VM) allocation and migration architecture which takes advantage of multi-layered neural architecture. The identification of the best VM using CS the information is passed to Support Vector Machine (SVM) and the trained support vectors are fed to the neural network. The simulation analysis conducted over 100 VMs had demonstrated that the proposed design proved to minimize the energy consumption by reducing the number of VM migrations while exhibiting least SLA violations. The comparative analysis against existing work demonstrated the percentage improvement of 6% in reducing number of VM migrations that reduced the energy consumption by 4% while exhibiting 2% reduction in SLA violations.

Published
2021-07-31
How to Cite
Satish Saini, K. S. (2021). Cuckoo Search inspired VM migration and allocation based on Neural Architecture. Design Engineering, 5733- 5747. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/3067
Section
Articles