CLOUD COMPUTING EFFECT EVALUATION AND VIRTUAL NETWORK OPTIMIZATION

Cloud computing paradigm has opened a new era of the Information and Communications Technology (ICT). Various research areas need to be considered with regards to cloud computing, such as scalability, availability, reliability(security), utilization, mobility, as well as cloud computing economics. The probabilistic demand models have been introduced in order to quantify the impact of cloud computing. The probabilistic demand models can clearly provide the effects of online migration to virtual machines. It can illustrate that the capacity of data centers can satisfy the demand on the service level with proportion of time. The performance of both the federation and non-federation can be evaluated through the service level concept. It triggers, therefore, a lot of research issues related to cloud computing. In this study, along with this trend, we consider federated cloud computing effect evaluation and virtual network optimization problems, which are the valuable research issues for cloud computing advance. In the federated cloud computing effect evaluation probabilistic demand models are investigated and the effects of cloud computing are clearly identified. For virtual network optimization mathematical programming approaches are studied and a new heuristic algorithm based on Simulated Annealing is proposed. From the computation results of the solution techniques our proposed heuristic algorithm provides improved performance than the previously developed heuristic algorithms called D-ViNE and R-ViNE.