Efficient Dynamic Routing in Elastic Optical Networks Based on Traffic Prediction and Bandwidth Reservation

In this paper, we study the problem of efficient dynamic routing in elastic optical networks. We propose a novel resource allocation strategy that benefits from the bandwidth reservation in advance based on the analysis of the traffic prediction. We also propose a mechanism for the strategy's partial implementation to provide its high flexibility and the possibility to control both the efficiency and processing time. Next, we perform numerical experiments on representative European network topology. Their aim is the selection of the strategy’s beneficial configuration as well as its performance evaluation. The results prove its high efficiency – the approach reduces the bandwidth blocking probability up to 0.85%, corresponding to 16.925 Tbps of more served traffic (compared to the allocation without traffic-aware bandwidth reservation).