IP Network Congested Link Location Algorithm Based on Bayesian Network

In the method of end-to-end (E2E) path performance active detection (network tomography), the link loss rate solution methods involve the inverse of linear equations, so its large computation complexity may lead to failure. This paper proposes a new congestion link location algorithm based Boolean model. First the nonsingular coefficient matrix of the linear system for solving prior probability is formed; then the prior probability of link congestion is calculated. Finally, based on Bayesian maximum a posteriori probability (MAP) estimator, the set of congested links can be located. The validity and accuracy of this algorithm is verified by experiments.