Evaluating the performance of optimal control based sensor deployment algorithms for realistic terrain model

Abstract

The extensive use of sensor networks in detection and surveillance systems has made optimal sensor deployment a key issue. A lot of sensor deployment techniques have been devised. While a number of mobile sensor networks consider the presence of obstacles and other terrain features in the region of interest, most static sensor networks either consider an idealistic basic terrain model or model the obstacles as straight lines. The aim of this paper is to extend upon a previously proposed optimal sensor deployment technique and incorporate into it a more realistic terrain model. This paper considers only those terrain features which can be modeled by convex polygons. Optimal sensor deployment together with a realistic terrain model makes sensor deployment more practical and robust. Simulation results show that the the number of sensors, their locations and the node density change in presence of obstacles.