SIAM Journal on Computing
Crosstalk-Aware Wavelength Assignment in Dynamic Wavelength-Routed Optical Networks
BROADNETS '04 Proceedings of the First International Conference on Broadband Networks
Translucent network design from a CapEx/OpEx perspective
Photonic Network Communications
DICONET NPOT: An Impairments Aware Tool for Planning and Managing Dynamic Optical Networks
Journal of Network and Systems Management
Novel strategies for sparse regenerator placement in translucent optical networks
Photonic Network Communications
| Hi-index | 0.00 |
Translucency in WDM networks appears as a trade-off between the low cost of full transparency and the high signal quality provided by full opacity. On the one hand, transparent networks undergo various transmission impairments due to optical components deployed in the network. On the other hand, opaque networks remain very expensive due to electrical 3R regeneration (Re-amplifying, Re-shaping, and Retiming), performed at each network node. Translucent networks use sparse regeneration strategy in order to improve the optical signal budget. In translucent network design, the objective is to judiciously choose the regeneration sites in order to establish a set of traffic demands with an admissible quality of transmission at a minimized network cost. We address the problem of translucent network design by proposing a novel heuristic for routing, wavelength assignment and regenerator placement. Our heuristic, called COR2P for Cross Optimization for RWA and Regenerator Placement, aims at minimizing both the number of required regenerators and the number of regeneration sites in the network. The originality of COR2P lies on a CapEx/OpEx perspective for network cost evaluation. Capital Expenditure refers to the network deployment cost while Operational Expenditure refers to the network management and maintenance cost. We introduce an original cost function that contributes to the optimization of CapEx/OpEx expenditures. In this paper, we investigate the impact of different parameters introduced in our heuristic and cost function, such as the ratio of sites chosen a priori for regeneration, and the limited size of regenerator pools installed at such nodes. Our simulation results outline that a tradeoff for CapEx and OpEx costs can be achieved by judiciously adjusting these parameters.