A heuristic analysis of resilience to multiple failures in GMPLS networks

Multiple failures can occur due to various causes, for example, natural disasters, epidemic outbreaks affecting software components, or intentional attacks. Such events can affect several network elements simultaneously, and the ensuing service disruption can be significant. This paper investigates new methods to evaluate and minimize the impact of such failures in terms of the number of affected connections. Two heuristic-based immunization strategies for improving network resilience are proposed. One strategy is built upon a new approach introduced in this paper, called the Observed Link Criticality. The other one uses the well-known concept of betweenness centrality. The strategies are evaluated through simulation. To that end, a large synthetic semi-complex topology is defined, representing a large GMPLS-based transport network, on which the normal operation of the network and the occurrence of large-scale failures are simulated.