Periodicity-and-Linear-Based Data Suppression Mechanism for WSN
Yanchao Song, Juan Luo, Chang Liu, Wenfeng He · 2015 IEEE Trustcom/BigDataSE/ISPA · 2015
For climate monitoring in wireless sensor networks (WSNs), there is a large amount of redundancy due to periodic and linear phenomena. In order to eliminate redundancy and save energy, current works utilize the linear phenomena to predict future data and suppress the sending of predictable data. If we predict future data based on both periodic phenomena and linear phenomena, we may predict and suppress more data. Based on this idea, in this paper, we proposed a Periodicity-and-Linear-Based data suppression mechanism (PLB). By counting the appearance frequency, appearance time and appearance duration of every linear pattern, PLB can calculate the Weighted Frequency of all linear patterns, which indicates the linear pattern's appearance possibility in the future. Based on the appearance possibility, PLB can choose a most probable linear pattern to predict the future data, which increases the prediction success rate and makes more data predictable. So PLB can suppress more data and save the energy of nodes. Simulation results show that: compared with previous works, the suppressed data amount of PLB has an obvious increase.