Approximation Algorithms for VirTUAL machine placement optimization in cloud environment

Abstract

We are living in a digital world where we are connected with various devices. It is an newlineera of cloud computing that is used by end-users from tiny pen drives to an expensive newlinecar. In the past few decades, the use of Cloud computing has not been limited to any newlinespecific domain but reach the boundary across the globe. Cloud computing has newlinevarious service models offered to various users. For example, google apps and newlineMicrosoft 365 are an example of the Software as Service model. Google app engine newlineand salesforce.com are an example of Platforms as a service model. The virtual newlinemachine, virtual private network, and cloud storage are a few examples of the newlineInfrastructure as a Service model. newlineIn this thesis, we have designed and developed new approximation algorithms and newlinesolution models to optimize virtual machine placement with balanced resource newlineutilization. All of these algorithms are experimentally evaluated on various newlinebenchmark datasets and real-time instances. The experimental result has shown very newlinepositive results in terms of balance resource utilization. newlineWe started with an in-depth literature study on Virtual Machine Placement newlineoptimization with various objectives in focus. Most of the existing techniques are newlinemissing multi-resource balance utilization during VM placement. They missed a newlineunified approach to utilize resources and were also limited to a single cloud newlineenvironment. A second observation is that algorithms and models were not tested newlineexperimentally on various benchmark instances and in a real-time cloud environment. newlineThe third major gap found in existing methods is that technique is very limited as a newline