TRAM-based VM handover with dynamic scheduling for improved QoS of cloud environment

Abstract

The generic view of cloud has seen changes with the entry of cloudlet which combines the servers located in any local area network that can be accessed through any wireless communication. Modern cloud computing has opened avenues for resource and service providers for deployment of their resources irrespective of location. The only issue is the choice of the relevant service provider for the cloud client. There are a number of scheduling algorithms discussed earlier. Resources have been grouped on the basis of various constraints. Yet, the methods are unable to achieve the required quality of service parameters. With a view to overcome the issue of previous methods, an efficient throughput-resource availability-makespan (TRAM)-based clustering, VM handover, with scheduling is discussed in this paper. The TRAM-based approach maintains a list of requests processed on each moment by different cloudlets and groups of the cloudlets according to the TRAM measure. The same TRAM measure is used for making decisions on the handover and scheduling of the request based on TRAM. The recommended approach produces auspicious results on the various qualities of service parameters and produces satisfactory results.