End-to-End Proportional Loss Differentiation in OBS Networks

Abstract

Proportional loss differentiation in OBS networks has recently received much attention since it allows to provide a quantitative differentiation between classes, so facilitating network operations and pricing to providers, but without the complexity involved by the absolute loss differentiation.Although edge-to-edge differentiation must be the ultimate aim, most of the methods in the literature are per-hop-based, and it is well-known that guaranteeing per-hop proportional loss does not guarantee end-to-end proportional loss.In this paper we modestly try to fill the gap which exists in this field, and so, we propose and analyze a new method to obtain edge-to-edge proportional loss differentiation for WDM-based OBS networks. This method is based on the idea of trunk and wavelength reservation, already used in circuit switched networks and deflection routing in OBS networks.Through extensive simulation experiments and analysis, we compare the performance of our mechanism to another mechanism proposed in the literature, based on using additional offset.KeywordsBurst SizeOptical Burst SwitchingCore NodeData BurstBurst Head PacketThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.