基于深度数据包检测(DPI)及R平方值的数据承载网规划方法的研究

Research of IP bearer networks planning method based on deep packet inspection(DPI) and Rsquare value

流量模型法通过深度数据包检测技术精准提取特定时间段内各种互联网用户的流量模型参数,根据市场发展目标,计算流量预测值;曲线拟合法通过对特定时间段内的流量数据作曲线拟合,选取R平方值接近1的函数对未来做预测。对两种方法得到的结果进行加权处理,得到相对准确的预测值,是数据城域网规划的目标。

The traffic model method accurately extracts the traffic model parameters of various Internet users in a specific time period by using deep packet inspection technology, and calculates the traffic prediction value according to the market development goal;the curve fitting method performs curve fitting on the flow data in a specific time period. A function with an Rsquared value close to 1 is selected to predict the future. Weighting the results obtained by the two methods to obtain relatively accurate prediction values is the goal of data metropolitan area network planning.