Towards a scalable design for survivable optical virtual private networks (O-VPNs)

The paper tackles the resource allocation problem for optical networks supporting virtual private networks (O-VPNs), in which working and spare capacities are allocated in the networks for satisfying a series of traffic matrices corresponding to each O-VPN. Based on the (M:N)ⁿ protection architecture defined in generalized multiprotocol label switching (GMPLS), we propose two novel integer linear program (ILP) models, namely ILP-I and ILP-II, aiming to initiate a graceful compromise between the capacity efficiency and computation time without losing the ability of addressing QoS requirements in each O-VPN. Experiment results show that, in terms of capacity efficiency, a significant improvement is achieved by ILP-I compared to ILP-II at the expense of higher computation time. Although ILP-II is outperformed by ILP-I, it can handle the situation with an arbitrary size of O-VPNs. We conclude that the proposed ILP models yield a scalable solution for capacity planning in survivable optical networks supporting O-VPNs based on the (M:N)ⁿ protection architecture.