论文部分内容阅读
基于通用网络流量模型,采用时间序列建模,提出了一种专用于无线传感器网络的卡尔曼流量预测算法KTP/WSN.通过NS2仿真采集流量数据,使用该算法对流量数据进行预测.结果表明,该算法可以提前一个甚至几个周期预测网络节点的拥塞情况,提前做好路由选择,实现路由自适应控制,预测值和原始值偏差很小.进一步进行流量预测可对网络的占空比、能耗等做到提前自适应控制.
Based on the general network traffic model, time series modeling is used to propose a Kalman traffic prediction algorithm KTP / WSN, which is dedicated to wireless sensor networks. The traffic data is collected by NS2 simulation and the traffic data is predicted using the proposed algorithm. The algorithm can predict the congestion of network nodes one or even several cycles in advance, make the routing in advance and realize the routing adaptive control, the deviation of the predicted value from the original value is very small.Furthermore, traffic forecasting can predict the network duty cycle, Consumption and other early adaptive control.