一种准确预测无线局域网业务量的时间序列模型

在对随机选取的实际环境中的多组无线局域网业务量数据的研究过程中，发现无线局域网业务量具有明显的多重季节性．利用差分和特定间隔抽样对原始数据处理，从而验证了无线局域网业务量的季节性．并提出了一个能够准确预测无线局域网业务量的时间序列模型：乘积型季节ARIMA(0,1,1)(0,1,1)₁₂模型，通过迭代计算，将此模型转化成一个MA模型，并且利用MA模型的性质对模型参数作出估计．利用差分方程法，对随机选取的一段无线局域网业务量进行了预测．结果表明，此模型可较好地对无线局域网的业务量进行短期预测，且提前10步预测的平均相对误差仅为0.0401．

A time series model for accurately predicting the WLAN traffic

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 (0.040 1) when the forecast steps are 10.