Optimal capacity placement for path restoration in STM or ATM mesh-survivable networks
IEEE/ACM Transactions on Networking (TON)
Optical networks: a practical perspective
Optical networks: a practical perspective
CNSR '06 Proceedings of the 4th Annual Communication Networks and Services Research Conference
Fault management in IP-over-WDM networks: WDM protection versus IP restoration
IEEE Journal on Selected Areas in Communications
A practical approach to operating survivable WDM networks
IEEE Journal on Selected Areas in Communications
Routing and wavelength assignment of scheduled lightpath demands
IEEE Journal on Selected Areas in Communications
A new model for allocating resources to scheduled lightpath demands
Computer Networks: The International Journal of Computer and Telecommunications Networking
Resource provisioning for survivable WDM networks under a sliding scheduled traffic model
Optical Switching and Networking
| Hi-index | 0.24 |
In this paper, we consider resource allocation under the scheduled traffic model, where the setup and teardown times of scheduled demands are known in advance. A number of integer linear program (ILP) solutions for this problem have been presented in the literature. Most of these ILPs minimize the number of wavelength links. A more appropriate metric, for wavelength routed, all-optical networks, is to minimize the congestion (the maximum number of lightpaths on a single fiber link) of the network. We present a new ILP formulation for routing and wavelength allocation, under the scheduled traffic model that minimizes the congestion of the network. We propose two levels of service, where idle backup resources can be used to carry low-priority traffic, under fault-free conditions. When a fault occurs, and resources for a backup path need to be reclaimed, any low-priority traffic on the affected channels is dropped. The results demonstrate that this can lead to significant improvements over single service level models. We also show that the complexity of our formulation (in terms of the number of integer variables) is lower, even compared to existing approaches, which only allow a single level of service. Therefore, we are able to generate optimal solutions for practical networks within a reasonable amount of time, whereas existing formulations often do not find the optimal solution even after 2h. Finally, we present a simple and fast heuristic that can quickly generate good solutions for much larger networks.