Dynamic programming: deterministic and stochastic models
Dynamic programming: deterministic and stochastic models
Optical networks: a practical perspective
Optical networks: a practical perspective
Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
BROADNETS '04 Proceedings of the First International Conference on Broadband Networks
Differentiated integrated QoS control in the optical Internet
IEEE Communications Magazine
IEEE Journal on Selected Areas in Communications
Service-specific resource allocation in WDM networks with quality constraints
IEEE Journal on Selected Areas in Communications
| Hi-index | 0.00 |
In this paper, we propose a service differentiation method for lightpath establishment in all-optical WDM networks. The proposed method derives the optimal lightpath establishment policy for each serve class based on a Markov Decision Process (MDP). The optimal policy considered here is to provide service differentiation for lightpath establishment and to utilize wavelengths effectively. Here, we focus on two cases for wavelength conversion capability; full-range wavelength conversion and limited-range wavelength conversion. For full-range wavelength conversion, the number of lightpaths for each class is considered as a state of MDP. On the other hand, for limited-range wavelength conversion, the service class of lightpath that has been established with each output wavelength is considered as the state. Based on these states, we derive the optimal policy for lightpath establishment. In both cases, we evaluate the performance of the proposed method with simulation. In numerical examples, we show that our proposed method can provide service differentiation and use wavelengths effectively regardless of wavelength conversion capability.