Sensor scheduling based on prediction

Aiming at the construction and reconstruction of dynamic cluster for target tracking with wireless sensor network, a sensor Scheduling target tracking-oriented algorithm based on prediction is proposed in this thesis, in which construction and reconstruction of dynamic cluster depends on prediction of nodes´ location. Selection of cluster head is achieved by weighted algorithm considering residual energy and geographic position or other parameters. The extended kalman filter is utilized to predict the trajectories of target in this paper and accuracy of target tracking is improved by cooperation in nodes.