A synthetic evaluation methodology based on neural networks theory for reliability indexes of communication networks

This treatise uses a neural network to evaluate a running-index system of reliability on communication networks. In the synthetic evaluation of the multiobjective, the calculation of the index weight is a key problem, as a neural network has neurocomputing, parallel distributed processing and self-adaptive learning characteristics, and can implement non-linear mapping. Making use of neural networks to evaluate the running-index system of reliability on communication networks is very effective and convenient. This treatise uses the Hamming neural network to evaluate the reliability indexes of communication networks. It illustrates the evaluation of the reliability of a network with a complicated topology by a practical example.