Finding Holes in Sensor Networks
The authors consider the problem of detecting holes in a sensor network. A hole occurs when several adjacent nodes in a sensor network fail, and is defined as the convex hull of the region containing the failed sensors. The presence of holes in sensor networks has important consequences on the information now and capacity across the network and on the perceptual coverage of the network. They present, analyze, and implement two distributed algorithms for detecting holes. In the first algorithm the neighbors of the failed nodes detect and compute the hole locally. The second algorithm considers the case of detecting a hole from a distance. They present implementation results of the first algorithm using a 30 node Mica Mote sensor network and simulation results of the second algorithm.