Research And Implementation Of Communication Traffic Forecast Based On The Seasonal Modeling Algorithm

Motivated by the real running and maintenance requirement of switching network and taking account of the telecommunication traffic data' characteristics and networking running indexes of quality, this paper deeply investigates the traffic data and time serial. Then the prediction technology of traffic data and the implementing methods are discussed and analyzed by using the original traffic data of telecommunication switches and adopting the seasonal models. The work of this paper includes:(1) Deeply investigating the characteristics of original traffic data of exchanger in telecommunication network , and summarizing the requirements of analysis and forecast.(2) Through analyzing the original traffic data, the function of analyzing and query is brought forward and investigated, then implemented. That enables the different people to directly master the information contained in the data and then to use flexible query. Moreover the users can analyze and inquire their interesting data by selecting aspects of different dimensionality.(3) Having accomplished the forecasting seasonal model based on Monte Carlo simulation, which is a new thought and algorithm for traffic forecast of telecommunications enterprise.The forecasting experiment is tested based on core thought of using the time queue. The input is monthly traffic data from the historical data of exchange. By taking the seasonal model algorithm to calculate and practice the traffic data, the results of the experiment show the changing rules of the concerned traffic. Further more, the future monthly traffic data is predicted according to the traffic characteristics in different areas.In order to better understand the predicting effect and analyse the reliability of forecasting model, the forecasting result using the way of the SEASONAL ARIMA is compared with other forecasting results using the models such as neural network forecasting under the same circumstances. The experiment result proves that the SEASONAL ARIMA has advantages over other models and can be applied to industry.