REASoN - Reliability and/or availability evaluation for sustainable networking

Sustainable computer networking approaches adapt network node power states dynamically in response to traffic demand. This scenario imposes new challenges to the way the network resilience is evaluated based on reliability and availability metrics. The traditional assessments of reliability and availability based on Markov models or the Cut-Set and Tie-Set techniques are based on static values and do not take into account the dynamic changes observed in an energy efficient network. We present REASoN, a method that extends and merges the Markov model and the Cut and Tie set technique in a way that allows an accurate evaluation of the network reliability and availability with dynamically adjustable power levels. A numerical evaluation of our method shows that even for a reduced network topology the impacts of energy efficient operations on reliability metrics are significant to a 1st decimal digit change in reliability.