The periodic data traffic modeling based on multiplicative seasonal ARIMA model

With the development of diverse applications in mobile network, the architecture of network becomes heterogeneous and complicated, which increases the complexity of network planning. Traffic modeling is a hot issue in network planning, and vast researches are committed to find a suitable model that can capture and reproduce various properties of a real trace. Besides, a good model should be able to predict the future network traffic efficiently. In this paper, we introduce a seasonal Autoregressive Integrated Moving Average model (SARIMA) to model the data traffic based on the property of periodicity in mobile network. With two actual traces from different areas, experimental results demonstrate that SARIMA model can effectively model and predict future data traffic.