Dynamic realization of miscellaneous profile services in elastic optical networks using spectrum partitioning

Abstract

Optical backbone networks are required to be highly adaptable in terms of bandwidth allocation, to accommodate the dynamic demands of emerging broadband wireless and fixed access networks. To address this requirement, rather than assigning a fixed bandwidth to each request, services are offered based on the requested bandwidth profile, which enables a more diverse range of services to be offered and more efficient usage of spectrum resources. To accommodate the service requests with miscellaneous bandwidth profiles, including minimum, average, and maximum spectrum slot requirements, as well as holding time, two innovative techniques have been developed. These schemes ensure that the allocated bandwidth meets the minimum requirement, does not exceed the maximum, and achieves the desired average bandwidth, considering the time-weighted average of the assigned spectrum slots over the holding period. These methods also utilize probabilistic spectrum partitioning, which enforces different probabilities to contributing spectrum partitions in a certain service realization. Employing this probabilistic spectrum partitioning along with profile-based routing, improves the chance of accommodating requests and consequently reduces request blocking probability. The results indicate that our algorithms can successfully realize the requested services by achieving a blocking probability of less than 0.07 for offered loads up to 1000 erlang, in the Deutsche Telekom network topology.