Risk assessment modelling for power optical backbone communication network based on MLFSR-oriented sampling Markov technique

In this paper, a hybrid multi-link failure statuses reduction (MLFSR)-oriented sampling Markov technique used for network service risk assessment is proposed. First, the link failure probability p is determined through link failure rate λ and repair rate μ. Then the link failure statuses can be reduced according to p, links number n and confidence level α which simplifies the failure state Markov model in assessing the possible service loss. Second, the sample size for each link failure status can be further reduced based on t bilateral percentile statistic theory in service loss estimation. Finally, the simulation and comparison are carried out for a practical communication network, and the results show that the network risk R has approximately linear relationship with link failure rate λ and mean repair time τ. This study is expected to have significant reference value for formulating risk-controlling measures to diminish communication network risk.