Probability model for data redundancy detection in sensor networks

  • Authors:
  • Suman Kumar;Seung-Jong Park

  • Affiliations:
  • Computer Science Department and Center for Computation and Technology, Louisiana State University, LA 70803, USA;(Correspd. E-mail: sjpark@csc.lsu.edu) Computer Science Department and Center for Computation and Technology, Louisiana State University, LA 70803, USA

  • Venue:
  • Mobile Information Systems - Advances in Mobile Communications and Computing
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Sensor networks are made of autonomous devices that are able to collect, store, process and share data with other devices. Large sensor networks are often redundant in the sense that the measurements of some nodes can be substituted by other nodes with a certain degree of confidence. This spatial correlation results in wastage of link bandwidth and energy. In this paper, a model for two associated Poisson processes, through which sensors are distributed in a plane, is derived. A probability condition is established for data redundancy among closely located sensor nodes. The model generates a spatial bivariate Poisson process whose parameters depend on the parameters of the two individual Poisson processes and on the distance between the associated points. The proposed model helps in building efficient algorithms for data dissemination in the sensor network. A numerical example is provided investigating the advantage of this model.