Minimization of the network availability upgrade cost with geodiverse routing for disaster resilience

Abstract

Telecommunication networks are a critical infrastructure of our society. Wide area backbone communication networks are based on optical networks, where each fiber has a very large capacity. These networks must offer high end-to-end availability and a high resilience to large-scale disasters. Routing with geodiversity can mitigate the impact of disasters but will result in longer paths, making it difficult to achieve the availability levels required by critical services. In this paper, we consider a given core optical network such that the current availability and the cost of upgrading it to a higher value are known for each network link (or edge). Then, the problem of selecting a set of edges to be upgraded at a minimum cost, while guaranteeing desired values of end-to-end availability and geodiversity, is considered and formulated as an arc based integer non-linear programming model. The non-linear constraints of the model are approximated and linearized, resulting in a new ILP based heuristic. A filtering procedure is proposed for decreasing (if possible) the cost and the number of upgraded edges of the solutions obtained by previously proposed greedy heuristics and also by the ILP based heuristic. The relative performance of the heuristics is evaluated using different geodiverse distances and end-to-end availability values in two reference core optical networks.