Kalman traffic prediction algorithm for WSN

Based on general model of network traffic, this paper uses time series as modeling tool, and then proposes an algorithm, named KTP/WSN (Kalman traffic prediction of WSN). Through NS2 (network simulator 2) experiment, the traffic simulation data were collected, and then were inputted to the algorithm. The results show that the algorithm can predict congestion status and determine the route in advance, and thus implement routing control adaptively. There is little variance between the prediction value and measured value. Further, the prediction value can be used in adaptive control to duty cycle and power.