Group-based Load Balancing Approach for Cloud Data Centre

Nowadays most of the IT organisations are using cloud to fulfil computation and storage requirements. Cloud computing provides various benefits such as high availability, fault tolerance, flexibility, reduce overheads and cost for users, scalability and on-demand services. Many of these depend upon proper load balancing. An efficient load balancing technique effectively utilises the resource and avoids uneven distribution among various jobs. In existing load balancing approaches, virtual machines are assigned to Virtual Machines (VM) without considering the requirements of the other jobs. This work introduces a new scheduling policy of VM allocation in cloud system which considers requirements of multiple cloudlets and selects the cloudlets for maximising the CPU utilisation. Furthermore, on comparing our results with other approaches, significant increase in performance has been observed and our technique outplays other techniques in all testing scenarios.