Tracking Mobile Targets Using Grid Sensor Networks
Ahmed M. Khedr · 2006
We have considered a sensor network where a lot of sensor nodes are spread in a grid like manner. T hese sensor nodes are capab le of storing data and thu s act as a separate dataset. T he entire network of these sensors act as a set of distrib u ted datasets. E ach of these datasets has its local temporal dataset along with spatial data and the geographical coordinates of a given ob ject or target. In this paper an algorithm is introdu ced that mines glob al temporal patterns from these datasets and resu lts in the discovery of linear or nonlinear trajectories of moving ob jects u nder su pervision. T he main ob jective here is to perform in-network aggregation b etween the data contained in the variou s datasets to discover glob al spatio-temporal patterns; the main constraint is that there shou ld b e minimal commu nication among the participating nodes. We present the algorithm and analy ze it in terms of the commu nication costs.