Research on Network Flow Evaluation Benchmark Model

Aiming at the rapid increase of network flow, especially the threat caused by abnormal flow to the network security and control, this paper puts forward a network flow evaluation benchmark model. Based on wavelet transformation and auto regressive moving average model, the purpose of the model is to offer reliable reference benchmark to the evaluation of network flow and help people establish the pre-warning mechanism of network flow. Experiments show that the flow the model fits has a good conformity with the actual network flow, which can reflect accurately the development trend and change degree of normal network flow, and can be used as the reference standard of flow evaluation.