# PoS(CENet2017)009

A New Fuzzy Neural Network Model and Its Application on Network Operating Situation Awareness

S. Liu, X. Li, X. Fan

Contribution: pdf

Abstract

In order to monitor the network in real time, scientifically analyze the network situation and predict its changing trend, we establish the Network Operating Situation Awareness (NOSA)
method based on new Fuzzy Neural Network (FNN)model, which has improved the well-known back propagation (BP) algorithm to accelerate the learning rate; besides, we use Wang-Mendel (WM) method to efficiently extract fuzzy rules. The s imulation results show that, in comparison with the traditional models, the new model features superiorities in convergence speed and learning accuracy, and gives a relatively accurate assessment of network operating situation. The new FNN model is a powerful method of NOSA, especially in complex environment, where there're massive uncertain and ambiguous data