BLITZ: wireless link quality estimation in the dark

  • Authors:
  • Michael Spuhler;Vincent Lenders;Domenico Giustiniano

  • Affiliations:
  • ETH Zürich, Switzerland;Armasuisse, Switzerland;ETH Zürich, Switzerland

  • Venue:
  • EWSN'13 Proceedings of the 10th European conference on Wireless Sensor Networks
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present BLITZ, a novel link quality estimator that relies on physical-layer synchronization errors to estimate the expected packet delivery ratio of wireless links. In contrast to all existing link quality estimators which estimate the packet delivery based on statistics from packets that are successfully decoded, our technique works even when packets at the receiver are not correctly received, i.e., when the synchronization fails. The core idea of BLITZ is to exploit information from chip errors in the received preamble of any transmitted direct sequence spread spectrum signals such as IEEE 802.15.4. Using extensive measurements over cable, wireless static and wireless mobile scenarios, we show that our proposed estimator outperforms existing estimators in terms of both accuracy and speed. Across diverse environmental conditions and the full range of possible link qualities, BLITZ provides packet delivery ratio estimates with an absolute error below six percent within just a few milliseconds.