International Journal of Engineering and Technology

Virtualization generally provides performance isolation between virtual machines running simultaneously on a single physical machine. But isolation does not happen to its fullest extent which may lead to performance interference. Due to this, the performance of applications running on one VM may get affected by running applications of co-existing VMs. This further leads to delay in execution time. As a result of this, throughput of cloud applications will be degraded. In order to deal with this situation, the proposed work has been presented which uses Particle Swarm Optimization based k-means++ algorithm for its implementation. The proposed work has been compared with some of the existing approaches and found to have better throughput and reduction in run time and cost.