Research on Network Traffic Forecasting Strategy Based on BP Neural Network

One major problem in the management of the current large networks is the complexity and the enormous amount of operations required to satisfy user demands while using resources efficiently. In this study, we propose a network traffic forecasting strategy based on BP neural network (BP-NTF). First, we analyse the characteristics of network traffic and establish traffic forecasting methods based on BP neural network, then modeling and forecasting the time series of network traffic data; Second, we construct three module, namely, data collection, data processing and traffic forecasting; Last, we use the strong memory and the learning ability of BP neural network to short-term forecast the network traffic .This model can provide a basis for network monitoring and management and has high application value and very wide meaning.