Robust virtual network embedding for mobile networks

Network virtualization which will, inter alia, allow for dynamic network sharing, has turned into one of the most prominent solutions that can efficiently deal with the upcoming increase of data demand in mobile networks. In order to allow multiple virtual network operators to share the same substrate infrastructure, the need for the development of efficient virtual network embedding algorithms and techniques is imperative. The main purpose of this paper is to provide a robust optimization framework for shortest path virtual network embedding, where the traffic demands as well as the user mobility are uncertain (stochastic) rather than deterministic parameters. The proposed algorithm takes into account the effect of the actual users´ mobility in order to allow efficient mapping for mobile wireless networks. The performance of the proposed algorithm is evaluated and compared with existing shortest path based virtual network embedding algorithms that are based on deterministic optimization problems, i.e., using nominal values on the demand. Numerical results reveal that the proposed algorithm can achieve virtual network embedding with adjustable robustness where the trade-off between conservatism and utilization of resources can be efficiently managed and controlled.