Efficient availability evaluation for transport backbone networks

The reliability of transport backbone networks is vital for the economy and security. Modern backbone networks use a mesh of fiber optic cables, which are, due to their ubiquitous deployment, prone to failures. The goal of this paper is to develop efficient computational methods for assessing the availability of such networks. We present both analytical and simulation approaches for this problem. Our analytical approach is based on cut set enumeration, while our simulation approach is based on Monte Carlo sampling techniques. To facilitate the computational process, we employ artificial intelligence methods based on genetic algorithms. We consider the expected loss of traffic (ELT) as the availability measure and show that ELT can be accurately estimated through an efficient procedure.