Traffic Model For Massive Games And Rallies

With the execution of the opening policy and the development of economy in China, more and more massive games and rallies were organized. Large group of subscribes bring big problem to the mobile network. But without the help of specified traffic model for massive games and rallies, it's difficult to deploy mobile stations and carriers within a reasonable budget.The specified traffic model for massive games and rallies required shout training sequence and less input data types. The classic forecasting model such as Time series forecasting method can't be used directly, because the source date is discrete. Regression analysis prediction method can't be used directly, because it's hard to give the exact number of the permanent subscribes in the target area. Modern forecasting model such as gray system or artificial neural networks was too complicated to be used by the non-professional.Here we give a new simply way, we divide the traffic in the target area into two separated part: the permanent part and the game part. Time series forecasting method can be used to forecast the permanent part, and the regression analysis prediction method can be used to forecast the game part. Finally, the two separated forecasting results are combined to get the forecasting value.The method is simply and effective, and it can be used in traffic forecasting for all kinds of massive games and rallies.