Efficient Simulation of Optical Ring Networks with Intermittent Transceiver Faults

Failures in the optical transmission equipment degrade the quality of service (QoS) provided by networks. In this paper, we consider intermittent faults and their impact on transceivers in optical network nodes. Our focus is on the service unavailability, i.e. the probability that the average capacity is below a specified guaranteed level. Since appropriate network models become huge and the unavailability should be very small, conventional analysis methodologies are usually not suitable. Here, we present an efficient adaptive importance sampling simulation technique. The change of measure which is crucial for the successful application of importance sampling is obtained using the ant colony optimization metaheuristic inspired by the foraging behavior of ants. That way, all necessary change-of-measure parameters automatically adapt to the model without requiring in-depth knowledge and pre-analyses which thus overcomes a major drawback in non-adaptive importance sampling. Numerical results demonstrate the accuracy and efficiency of the simulation technique.