The Study Of Energy Aware Vitual Network Embbeding

Virtual Network Embedding is a key technology to solve the future development of the Internet.The technology enables multiple heterogeneous virtual networks coexist in the same physical network,sharing the underlying resources.Most of the traditional virtual network embedding algorithms are aimed at the situation of high traffic load.When the traffic load is low,the virtual network embedding may cause the substrate resources not to be fully utilized,resulting in waste of network energy consumption.Energy aware virtual network embedding algorithm,through the consolidation of the substrate network resources,can ensure solving the energy waste of virtual network embedding process without affecting the communication quality.In this thesis,the problem of energy aware virtual network embedding based on node resource consolidation is proposed.Firstly,the corresponding network model and energy consumption model are established based on the analysis of the existing algorithms.A new substrate node ranking method is proposed for node mapping to realize the consolidation of the physical resources.On this basis,the corresponding energy aware virtual network embedding algorithm(EA-VNE-A)is proposed.The heuristic algorithm EA-VNE-A is compared with the energy-aware algorithm EA-VNE and the cost-based algorithm CostGRASP.The results show that EA-VNE-A has a good energy-saving effect.In the energy aware virtual network embedding based on splittable flow,the mapping scheme is further optimized and the corresponding network model and energy consumption model are given.When the substrate resources are integrated,the link resources are split to further improve the utilization rate of the substrate resources.The corresponding Mixed Integer Programming is given and the algorithm of energy aware virtual network embedding based on splittable flow is proposed.Comparing the VNE-EA-SF with the energy-aware algorithm VNE-EA and the cost-based CostVNE,the results show that the VNE-EA has better energy-saving performance.