Impairment-aware routing and wavelength assignment in translucent networks: State of the art

In the last 15 years, numerous investigations by both academia and industry have been carried out in the field of all-optical WDM networks' design. In all-optical - or transparent - WDM networks, data is transmitted from its source to its destination in optical form, switching/routing operations being performed in the optical domain without undergoing any optical-to-electrical conversion. Optical transparency may considerably reduce network infrastructures' cost and extend the range of services offered by the carriers. Designing an all-optical network consists of assigning to each traffic demand an endto- end optical circuit, also called "lightpath." In such networks, the problem of routing and wavelength assignment (RWA) aims to find an adequate route and an adequate wavelength for each traffic demand subject to the wavelength continuity constraint and limited network resources. The feasibility of the obtained lightpaths in terms of admissible quality of transmission (QoT) presents another difficulty. Indeed, according to the state of technology, various physical impairments degrade the quality of the optical signal along its propagation. Optical fibers and optical amplifiers as well as optical switching/routing nodes impact on end-to-end QoT. In this context only translucent networks are achievable, for instance, at a pan-European or pan-American scale. A translucent network uses electrical regenerators at intermediate nodes only when it is necessary to improve the signal budget. The cost of a network is roughly proportional to average number of input/output ports of a node. Knowing that today an optical port is five times less expensive than an electrical one, sparse regeneration allows translucent WDM networks to meet the QoT requirements and achieve performance measures close to those obtained by fully opaque networks at much lower cost. In this article we propose a state of the art in the field of impairment-aware RWA (IA-RWA), starting from the case of predictable traffic demands to the open problem of stochastic traffic demands. An economic analysis of the IA-RWA problem is proposed to justify the concept of translucent networks. The case of multi-domain lightpath establishment is also considered. Several examples of still open problems are mentioned in the article. Most of the concepts and results presented in this article refer to the FP7 DICONET European project in which the authors are involved.