Enhancing the Energy Efficiency of Virtual Network Embedding in Cloud

Abstract

Network Virtualization is recognized as a key technology for the future internet. Energy-efficiency is one among the most challenges in future networking environments. Network virtualization has caught the attention of the many researchers in recent years. It facilitates the method of making many virtual networks over one physical network. In existing system, if any one of the machine will get broken means that we are able to simply transfer all the information corresponding to that machine to a different machine. Here we tend to concentrated on only data we don’t take into account network efficiency, cost and power consumption.to overcome this problem we move to proposed model. During this paper, we tend to propose an energy economical virtual network embedding (EEVNE) approach for cloud computing networks, wherever power savings are introduced by consolidating resources within the network and data centers. We model our approach in an imp over WDM network using mixed integer linear programming (MILP). The performance of the EEVNE approach is compared with 2 approaches from the literature: the bandwidth cost approach (CostVNE) and also the energy aware approach (VNE-EA). The CostVNE approach optimizes the utilization of available bandwidth, whereas the VNE-EA approach minimizes the ability consumption by reducing the amount of activated nodes and links while not taking into consideration the granular power consumption of the information centers and also the completely different network devices. The results show that the EEVNE model achieves a most power saving of 60% (average 20%) compared to the CostVNE model below an energy inefficient data center power profile. we tend to develop a heuristic, real-time energy optimized VNE (REOViNE), with power savings approaching those of the EEVNE model. we tend to additionally compare the various approaches adopting an energy efficient data center power profile.