Inference Method for Lossy Links Based on Prior Probability
Wandong Cai · Jisuanji gongcheng · 2010
This paper studies identification of lossiest links based on Boolean Network Tomography. It addresses this problem of identification of lossiest links to the most probable explanation in Bayes estimation. A method using link state prior probability is presented to inferred lossiest network links, the presented method computes posteriori conditional probability of each candidate link assignments and chooses the assignment with maximum probability. Theoretic analysis proves that the method is feasible, and simulation results show that the method achieves accuracy higher than previously proposed heuristic methods.