Loss inference in wireless sensor networks based on data aggregation

  • Authors:
  • Gregory Hartl;Baochun Li

  • Affiliations:
  • University of Toronto, Toronto, Ontario;University of Toronto, Toronto, Ontario

  • Venue:
  • Proceedings of the 3rd international symposium on Information processing in sensor networks
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we consider the problem of inferring per node loss rates from passive end-to-end measurements in wireless sensor networks. Specifically, we consider the case of inferring loss rates during the aggregation of data from a collection of sensor nodes to a sink node. Previous work has studied the general problem of network inference, which considers the cases of inferring link-based metrics in wireline networks. We show how to adapt previous work on network inference so that loss rates in wireless sensor networks may be inferred as well. This includes (1) considering the per-node, instead of per-link, loss rates; and (2) taking into account the unique characteristics of wireless sensor networks. We formulate the problem as a Maximum-Likelihood Estimation (MLE) problem and show how it can be efficiently solved using the Expectation-Maximization (EM) algorithm. The results of the inference procedure may then be utilized in various ways to effectively streamline the data collection process. Finally, we validate our analysis through simulations.