Abstract: During the process of studing much random selected real environmental WLAN traffic, the multiplicative seasonal
property of WLAN traffic has been discovered. By the use of differencing and specific sampling of the orginal data sequence,
the seasonal property is verified in this paper. A time series model is given which can accurately predict the WLAN traffic:
Multiplicative Seansonal ARIMA Model (0,1,1)(0,1,1)₁₂. After serveral iterative computations, the model is transformed into
an MA model and its parameters have been estimated using the character of the MA model. A prediction of the random selected
WLAN traffic has been finished by the differencing function. The result of the prediction shows that the proposed model can
short-term forecast the WLAN traffic and obtain a better result with an average relative error of only 0.0401 when the
forecast steps are 10.

Key words: WLAN, ARIMA, forecasting of network traffic, time series