In modern optical networks, infrastructure management is faced with the challenge of using expensive equipment and communication resources as efficiently as possible. This now includes keeping power consumption costs at a minimum and using the available optical links in a balanced way, in addition to the traditional goals of providing the best possible performance to the end customers while meeting their quality requirements. Accordingly, this paper presents a heuristic single-step lightpath routing and wavelength assignment algorithm, handling online dynamic connection requests within a fully distributed network control plane. By using shortest path routing, the presented scheme determines the best compromise solution between the users' and carrier's objectives. The former can be mainly expressed in terms of connection QoS requirements, while the latter comprises network engineering (distributing the load in order to achieve near-optimum resource usage) and containing energy consumption. This approach is able to find, in a polynomial computing time, a multiobjective optimization solution that maximizes the carriers' return of investment and supports a high number of users' request while drastically reducing the network operational expenditures, as extensively demonstrated through a simulation.

Peer Reviewed