Energy efficient aggregate monitoring for sensor fields with time-varying statistics
Esmaeil Biazar Amlashi, Ahmad Khonsari, Samaneh Heidari, Ronak Amani · International Journal of Sensor Networks · 2016
Monitoring in wireless sensor networks (WSNs) has been considered an issue of important value in research community. Reactive monitoring mechanisms are important in WSNs because of their low energy consumption. The energy efficiency of such mechanisms is relying on setting proper thresholds that act as the parameters of distributed local filters. In this paper, we address how to appropriately design local filters for reactive aggregate monitoring scenarios in random sensor fields with dynamic statistics. Since previous optimal local filter designs are proposed for stationary fields, they cannot work efficiently in the presence of statistical variations in the sensor field. We propose dynamic sigmoid approximated threshold assignment (D-SATA) algorithm, which is a distributed algorithm that adapts the optimal local filter design to the variation of statistics of sensor field. Simulation experiments confirm that the offered monitoring mechanism of D-SATA is efficient in terms of overhead burden to the network.