Dynamic power management of WSN based on data stream grey forecasting

Due to the characteristics of limited computing capabilities and small storage of WSN node, a dynamic power management (DPM) method that can realize data stream forecasting based on the grey theory was proposed. Two working modes were established in terms of the different data receiving/sending cycle status of the node, combining the features that grey forecasting has strong adaptive ability and needs few samples and computation work, the model for forecasting the data stream between node and converged node was established to realize the dynamic switch of working modes and reduce the energy consumption on the communication between the WSN nodes. The grey forecasting of data stream on the basis of only 4 sample data is able to obtain the forecasting result of MSE=1.21%. The simulation analysis of nodes´ energy consumption indicates that this DPM method can save 67.9% of energy.