Sensor placement based on Delaunay triangulation for complete confident information coverage in an area with obstacles

This paper studies the sensor placement problem for ensuring complete coverage in an area with obstacles. Instead of using the simplistic disk coverage model, we adopt our recently proposed confident information coverage model for field attribute monitoring applications. We propose a node placement algorithm based on iterative Delaunay triangulation, which is to first obtain Delaunay triangles for some initial seed nodes. Among all Delaunay triangles, we propose algorithms to find a valid one yet with the largest coverage hole for placing a new node. The Delaunay triangulation process is then repeated, until all the Delaunday triangles can be completely covered. Simulation results show that our algorithm has comparable performance in terms of the number of placed nodes, compared with a peer algorithm based on a grid approach to discretize the continuous field. However, our algorithm can truly achieve complete coverage yet with significantly smaller computation time.