Optimizing Deadline-Driven Bulk-Data Transfer to Revitalize Spectrum Fragments in EONs [Invited].

Wei Lu,Zuqing Zhu,Biswanath Mukherjee
DOI: https://doi.org/10.1364/jocn.7.00b173
IF: 4.508
2015-01-01
Journal of Optical Communications and Networking
Abstract:This paper studies how to optimize deadline-driven bulk-data transfer to recycle spectrum fragments in elastic optical networks (EONs). We first formulate a mixed integer linear program (MILP) for offline optimization with one of two objectives: 1) maximize the average data-transfer percentage of data-oriented requests (DO-Rs), or 2) minimize the incompletable data transfers among the DO-Rs. Then, we propose two dynamic provisioning algorithms. One addresses the first objective by setting a threshold on the minimum transferred data per time slot, and the other considers both admission control and blocking-aware routing and spectrum allocation to optimize the service provisioning according to the second objective. Simulation results indicate that the proposed algorithms can effectively optimize the DO-Rs' data transfers according to their objectives, revitalize the spectrum fragments, and improve the network spectrumutilization significantly.
What problem does this paper attempt to address?