On the optimal random deployment of wireless sensor networks in non-homogeneous scenarios

  • Authors:

  • J. Vales-Alonso;F. J. Parrado-GarcíA;P. LóPez-Matencio;J. J. Alcaraz;F. J. GonzáLez-CastañO

  • Affiliations:

  • Department of Information Technologies and Communications, Technical University of Cartagena, Spain;Department of Information Technologies and Communications, Technical University of Cartagena, Spain;Department of Information Technologies and Communications, Technical University of Cartagena, Spain;Department of Information Technologies and Communications, Technical University of Cartagena, Spain;AtlantTIC, University of Vigo, Spain

  • Venue:
  • Ad Hoc Networks
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Random scattering of WSNs is needed in many practical cases due to the large scale of the network required or to the inaccessibility of the terrain. However several important features of deployments of this type have been neglected due to their analytical complexity. Node placement must guarantee correct operation: if nodes are too separated many would be isolated and data would not reach the sinks. Besides, if the nodes are too close, the area covered would be small and little information would be retrieved. Moreover, the target area cannot be considered homogeneous since in real-life situations some zones are more important than others. This paper addresses these constraints by proposing and solving an optimization problem which maximizes network sensing coverage. In our model several clusters of nodes are spread over the target area following Gaussian random distributions, and the goal is to decide the optimal launch point and the dispersion for each cluster. This corresponds to real situations where clusters are dropped in an airborne launch in which dispersion is controlled by the release altitude. The problem is solved by considering iterative steps where single cluster deployments are addressed. Several tests validate our approach and indicate that our method outperforms previous approaches, especially in deployments with a low number of nodes, which are more challenging from the optimization perspective.