Abstract
In this paper, we present a scalable framework for regenerator placement (RP), routing, modulation level selection, and spectrum allocation (RMSA) jointly in translucent elastic optical networks. We introduce first a path-based mathematical formulation to solve RMSA problem given a designated set of regenerator site locations. In the preceding RP phase, the MILP model was invoked repeatedly to find the optimum sequence of regenerator site endowment. The scalability issue of MILP model is tackled by partitioning unserved demands into small groups and solve the proposed MILP model in recursive manner. In contrast to previous works that suggested random, volume, and hop-length-based one-pass demand ordering, we devise a novel adaptive routing and impact-based demand grouping and ordering scheme that outperforms the conventional random or resource consumption-based ones. We show that the proposed recursive MILP yields solutions to RMSA problem closely fitted to the ideal nonrecursive MILP counterpart under a wide range of traffic load variation even when the group size is reduced to one. In order to extend further the scalability of the recursive MILP, we develop heuristic algorithms fueled by the notion of availability-based spectrum assignment, impact-based path selection, and demand grouping and ordering. Numerical simulation verifies that the proposed heuristic obtains nearly the same performance as the recursive MILP but in very shorter running times. This enabled us to precisely analyze the behavior of the Deutsche Telecom network operating in real-like fiber passband setting under heavy traffic load. We show that the attainable network throughput exceeds 100 Tb/s.
© 2016 IEEE
PDF Article
More Like This
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription