Efficient Technique for Allocation of Processing Elements to Virtual Machines in Cloud Environment

Cloud Computing is an efficient way of providing services and resources to the users through Internet. It is an efficient way of sharing computing resources between a number of users simultaneously in an economic way. It is based on pay per use model, i.e., the user will have to pay only for the services or resources consumed by him / her. Now a day, it has been widely used in a number of areas. It can be characterized by the following features like distributed, heterogeneous, flexible, scalable, location independent, on demand self service and ubiquitous network access. The most attractive feature of cloud computing is that it can be accessed through the underlying physical infrastructure. It means that the users need not to build new infrastructure to access the cloud services and resources. The operation of a cloud environment is totally virtualized, i.e., the user of the cloud system does not have any information about the underlying physical infrastructures, servers where the applications and services are miming and the location of the physical resources and the data. In such systems, there is a great need to provision the operation of the system for proper load balancing. For this purpose, the concept of scheduling has been incorporated in cloud systems. Scheduling in cloud systems is performed at two levels, viz. at the host level (Allocation of PE's to VM's, called as VM Scheduling) and another at the user level (Allocation of Cloudlets to VM for execution, commonly called as Cloudlet Scheduling). In this paper, we are proposing a new algorithm for efficient allocation of PE's to VM's, so that the utilization of the physical resources provided by the cloud service provider can be improved to a certain extent. For this, we are using an optimization algorithm of the operational research domain called as Ant Colony Optimization. Our proposed algorithm can be considered as an improved form of time shared VM scheduling.