Realistic data transfer scheduling with uncertainty

On the dynamics of valley times and its application to bulk-transfer scheduling

“…In this sense, we are more interested in the typical shapes of the demand curves than in the low-grain description of bandwidth/usage time series, which has been extensively studied in other works-such as [45,42,43], where the authors focused on the prediction of traffic in the scale of a few minutes, with applications to cellular communications using neural networks and exploiting the number, distribution and class of users in a base station or the saturation of the access and mobility management functions in 5G networks, respectively. Moreover, we have also tried to minimize assumptions and information needs to reduce dependencies on the availability of information apart from traffic time series.…”