Service Scheduling and Resource Allocation for Big Data Transfer in Elastic Optical Network

A major class of traffic demands entails the transfer of known volumes of data before a deadline specified by the demands, while allowing slow resource allocation. For the service provisioning of these demands on an elastic optical network architecture, on which few studies have been carried out thus far, a model must handle both the time and the severe spectrum defragmentation resulting from the resource allocation. In this paper, a model is introduced that by harnessing the characteristic of these demands, targets efficient data transfer while meeting the demands requirements. It consists of two complementary models: a mathematical optimization model with a high degree of freedom as the core for resource allocation and admission, and a heuristic algorithm for service scheduling and service update. The results of the numerical examinations show the effectiveness of the presented model.