Boundary Detection and Area Estimation of an Event Region in Wireless Sensor Networks using Digital-Circles

In wireless sensor networks with large number of sensor nodes, deployed to monitor a given region, it is often required to detect and localize a critical event, like forest fire, or chemical pollution etc., in real time. The challenge is to select a minimal set of affected nodes to report so that the desired accuracy of area estimation can be achieved keeping the latency low. So far, in the literature, it has been assumed that the area enclosed by the points of location of all affected nodes, necessarily defines the event area. But in reality, a sensor node senses not just a point, but a region, ideally a circular one, determined by its sensing radius. Considering this realistic model of sensing, in this paper, we follow a digital geometry based approach to identify a minimal set of boundary nodes to report, and hence to estimate the affected area, assuming that each sensor node senses a circular area approximated by a digital circle. Simulation studies show that even for a sparse network the proposed technique may estimate the area with high accuracy using simple in-node processing.