دانلود رایگان مقاله روش کنترل بازخورد برای انرژی کارآمد تعبیه شبکه مجازی

Abstract

Network virtualization is an enabler for intelligent energy-aware network deployment. The existing research usually searches the subset of resources in the whole substrate network passively for the virtual networks (VNs), where resource consolidation achieves the minimization of energy consumption by switching off or hibernating as many network nodes and interfaces as possible. However, the stable active resources for accommodating the VNs help enhance the number of the hibernated nodes and links, which can reduce the energy consumption. A novel method for energy efficient virtual network embedding (EEVNE) is proposed in this paper, which controls the mappable area of the substrate network actively and can find the minimal consolidation of the network resources. In our proposed method, a controller, a check device and an actuator are designed for finding the stable consolidated subset of the substrate resources. Besides, two feedback-control-based EEVNE algorithms are devised to minimize the energy consumption of the substrate network for embedding VNs. Simulation results show that our algorithms significantly save greater energy than the existing algorithms.

6. Conclusions

Due to the dynamic characteristics of VNE, the active resources of SN change frequently, more quantities of the nodes and links are activated and more energy consumptions are produced. In this pa- per, we presented a novel feedback control approach for EEVNE. The stable consolidated subset of the substrate resources can be found for current VNs. Two feedback-control-based algorithms were pre- sented, which can increase the number of the hibernated links and nodes, as a result, the energy consumption can be reduced remark- 912 ably. Our algorithms in both theoretical analysis and simulations have shown their superiorities. The experiments demonstrate that a mini- mal subset of the substrate nodes and links for VNs can be found. The number of the hibernated nodes and links is enhanced, and the en- ergy consumption is reduced significantly in the non-saturated sce- narios.