Abstract:  A combination forecasting approach for Quality of Service based on Radial Basis Function neural network (RBF) is proposed, which uses time series model to establish linear and nonlinear forecasting models, and chooses the optimal model, then establishes different size sliding window dimension gray filling forecasting model according to the data characteristics. The forecasting results of these two models are passed into the RBF training model as the input source, and then begin to forecast. The experimental results show that our approach is better than existing models and improves the accuracy of prediction.

Key words: QoS, combination forecasting, RBF neural network, grey forecasting, time series