On the influence of the underlying network topology on optical telecommunication network availability under shared risk link group failures

Network availability is of paramount importance in optical telecommunication networks. Their rising connectivity and consequently their availability is compromised by link and node failures, usually due to physical force (e.g. digging, earthquake or fire). Single link failures can in turn cause multiple failures in case a failure hits a shared risk link group (SRLG), which is a group of logically distinct links sharing a common physical resource, be it a cable or a conduit. The number and length of SRLGs, as well as the characteristics of the underlying physical topology can significantly affect network availability. Especially, the physical topology can be represented by realistic synthetic graphs which are created by numerous geographic graph generators. This work describes the implementation and usage of six different physical topology models (Random Geometric, Gabriel, Relative Neighborhood, K-Nearest Neighbor, Waxman and Spatial Barabási-Albert) for investigation of the influence of the underlying topology on the optical telecommunication network availability. Network availability is estimated using Monte Carlo simulations based on a model of optical telecommunication network implemented by network simulator ns-3. Scenarios utilizing six topology models both in absence and presence of SRLGs are studied, and the optical network availability sensitivity to the underlying physical network topology is presented as the main result.