A 3d-localization and terrain modeling technique for wireless sensor networks

  • Authors:

  • Hady S. Abdelsalam;Stephan Olariu

  • Affiliations:

  • Old Dominion University, Norfolk, VA, USA;Old Dominion University, Norfolk, VA, USA

  • Venue:
  • Proceedings of the 2nd ACM international workshop on Foundations of wireless ad hoc and sensor networking and computing
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Although sensor networks are usually deployed in complex 3D terrains, the majority of localization techniques proposed in the literature are designed assuming 2D deployments. Furthermore, in general it is not easy to extend these techniques to 3D deployments. With minimal communication and computation overhead, we present a 3D-localization technique that comes with a terrain modeling capability that makes the technique more interesting. Sensors are localized in two steps. First, we determine the horizontal plane within which the sensor lies. Then, we project the nearest three anchors to the sensor onto the horizontal plane determined in the first step. Finally, we use RSSI-based distance measurements and trilateration techniques to fully localize the sensor. After localization, we use Delaunay triangulation, to build a mesh that models the terrain of the deployment area. In addition to the terrain modeling capability, the proposed technique has several attractive features, such as scalability and independence from the underlying network topology. Simulation results demonstrate that the proposed approach can achieve high accuracy in sensor localization and terrain modeling under simple and moderate 3D deployments. Intuitively, the accuracy of the modeled terrain increases with network density as in such cases there will be more vertices available to construct a more accurate mesh.