A data sample algorithm applied to wireless sensor network with disruptive connections

This paper presents a data sample algorithm applied to wireless sensor network applications with disruptive connections. Additionally, it defines a model for delay-tolerant sensor network where drop strategies are applied to improve the phenomenon coverage in an application that monitors the forest temperature incidence for wildlife observation. The environmental application model comprises: i) Phenomenon generation based on a Gaussian random field along with the Matern covariance model; ii) Sensing nodes deployment based on simple sequential inhibition process with a mobile sink node following a random walk process; iii) Data collection and processing based on a data-aware drop strategy; and iv) Phenomenon reconstruction based on simple kriging interpolation. This research employed the data-aware drop strategy and compared it with the others, reported in the literature. Besides the satisfactory application of this model, the results show that the performance of data-aware drop strategy is twice better than conventional ones in all evaluated scenarios.