The cloud service providers often fail to cope with the
increasing costs for rendering services and to mitigate the
increasing expenses; in some cases, they have to violate
the SLA. Though many models for managing servers and
deploying VMs is prevalent now-a-days, energy efficient server
management model is scarce and many of them are not
suitable for practical deployment. Again most of the models
do not take the bandwidth cost into account. Network devices
contribute largely to the energy expense of data centers. Communication aware VM scheduling can reduce the operational
energy expenditures and bandwidth requirements. We predict
resource utilization for the incoming requests using LPF [12].
Transferring data between two nodes connected directly to the
same switch tends to use less energy than that of the nodes
connected indirectly [13]. Our proposed algorithm prioritizes
data transfers among those nodes which are on the same
switch, in support of energy efficiency and delay minimization