Virtual optical network embedding of time-varying traffic in elastic optical networks

Abstract

Network virtualization can enhance the flexibility of Internet infrastructure resources and the scalability of the substrate elastic optical networks (EONs) through the coexistence of multiple virtual optical networks supporting different services. Traditional virtual optical network embedding mostly assumed fixed embedding, that is, once a virtual optical network (VON) is embedded, the network resources allocated to it are almost unchanged during its lifetime. However, the actual traffic demand on a VON usually changes over time. To address the problems of high bandwidth blocking probability and low spectrum utilization caused by the mismatch between the resources allocated and the required resources during the virtual optical network embedding of the time-varying traffic in the elastic optical networks (EONs), we propose a virtual optical network embedding algorithm of the time-varying traffic (TVT-VONE). Firstly, in the node embedding stage, a node weight value measurement is designed based on the attributes of virtual nodes and physical nodes. Secondly, in the virtual optical link embedding stage, a measurement formula for path spectrum resource richness is designed to decrease link embedding failure due to insufficient spectrum resources on the path. Finally, in order to adapt to the characteristics of dynamic changes in bandwidth requirements of time-varying traffic, we design a resource-aware virtual network reconfiguration strategy when the spectrum resources of traffic required do not match the resources allocated to the changed traffic. Simulation results show that the proposed algorithm can effectively reduce the bandwidth blocking probability of VON requests and improve spectrum utilization of EONs.