Survivable Virtual Network Embedding Algorithm Research

In order to deal with the problem of network "bloated" and provide experimental environment for network innovation architecture, network virtualization technology has been paid more and more attention. In order to realize the efficient virtualization, the corresponding virtual network mapping algorithm is needed. But in the current stage of the virtual network embedding algorithm, most researches are from the point of how to increase the utilization rate of the substrate network and reduce the time complexity of the algorithm and so on, and the research on the survivability of virtual network is very few. In this paper, we design a virtual network embedding algorithm based on the survivability of virtual network. First of all, it is the research of the global survivability, that is,when the substrate network fails, the number of the virtual requests which are embedded through the survivability algorithm is smaller than the other algorithms. In this paper, we design a virtual network embedding algorithm based on k-core theory, using the k-core theory of complex networks. This algorithm can decompose the complex virtual network and transform it into a network composed of shell structure, and then embed it onto the substrate network. At the same time, in the virtual network mapping algorithm based on k-core theory, the idea of load balancing is introduced. In the link mapping phase, the virtual link and physical link are processed in a slicing process, so that the virtual request can be embedded to the substrate network. The simulation results show that the virtual network embedding algorithm based on k-core theory not only improves the survivability of the virtual network from the overall situation, but also can improve the income of the substrate resources of the unit. Then from the perspective of the survivability of only one virtual network, that is when the substrate network fails, the virtual network can still work, we design a virtual network embedding algorithm based on a link self-backup. This algorithm is based on the minimum depth spanning tree algorithm, which will extend the link and bandwidth of the virtual request, and then embed it onto substrate network. Finally, the simulation results show that the virtual network embedding algorithm based on link self-backup can effectively reduce the backup resources and improve the revenue.