Virtual network embedding with adaptive modulation in flexi-grid networks

Network virtualization has been proposed as a promising method to mitigate the ossification of the Internet by allowing multiple heterogeneous virtual networks (VNs) to coexist on a shared substrate network. One of the major challenges in this method is the VN embedding (VNE) problem of how to map efficiently the virtual nodes and links onto the substrate network considering constraints associated with different substrate networks. This paper aims to solve the VNE problem with geographical constraints in the context of flexi-grid optical networks where modulation modes can be selected optimally. We provide an integer linear programming (ILP) formulation for the problem with the objective function of minimizing the embedding cost of an arriving VN. To achieve scalability, we also propose three polynomial-time heuristic algorithms where virtual links are embedded sequentially by three different sequences, respectively. We find that the sequence considering the bandwidth requirements of the virtual links outperforms the others. Such a sequence leads to a cost-effective VNE solution in terms of spectrum resource usage, which aims to optimize modulation modes and transmission distances of the virtual links that have high bandwidth requirements. Numerical results show that the heuristic algorithm with the sequence considering the bandwidth requirements performs closely to the ILP for a small size network, and we also demonstrate its applicability to larger networks.