A Biologically-Inspired Clustering Algorithm Dependent on Spatial Data on Sensor Networks

Sensor networks in environmental monitoring applications aim to provide scientists with a useful spatiotemporal representation of the observed phenomena. This helps to deepen their understanding of the environmental signals that cover large geographic areas. In this paper, the spatial aspect of this data handling requirement is met by creating clusters in a sensor network based on the rate of change of an oceanographic signal with respect to space. Inspiration was drawn from quorum sensing, a biological process that is carried out within communities of bacterial cells. In this system, global behaviour emerges from small-scale local events and this is an ideal characteristic of sensor networks. A spatial data model that showed the variation of water height as waves flow from the sea to the shore was used with real temporal data to test the algorithm. The paper demonstrates the control the user has over the sensitivity of the algorithm to the data variation and the energy consumption of the nodes while they run the algorithm.