Location-constrained survivable network virtualization

Network virtualization is regarded as a promising solution to the revolution of the current Internet. With virtualization, disruptive technologies can be easily deployed over the logic (service) network and transparently mapped to the physical network via the network embedding process. In this work, we address the survivability issue in the virtualization context via resolving the survivable network embedding problem. Particularly, we investigate the impact of the location-awareness and study the resulted problem, namely, location-constrained survivable network embedding (LSNE) problem. For the first time, we present an Integer Linear Programming (ILP) model to achieve joint optimal resource allocation for both the working and backup demand. For large-scale problems, we propose an efficient heuristic algorithm, which is shown to be close to the optimal results from the ILP.