Worst-case drift detection of sensor networks: performances and algorithms - IMT Atlantique Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Worst-case drift detection of sensor networks: performances and algorithms

Résumé

The goal of this paper is to derive algorithms that are able to detect unreliable/drifted sensors, in a sensor network. To recover the state of each sensor, we restrict ourselves to a particular decoder, inspired by graph partitioning problems. We provide necessary and sufficient conditions over the measurements such that the decoder perfectly recovers each sensor binary state. The outputs of the decoder can be computed using a dynamic programming approach. One challenging part of this approach is the complexity of the dynamic programming equation. Indeed, the resolution time will increase exponentially with the number of sensors. Therefore, we propose an efficient heuristic method that approximately solves the problem. We study the performance of our algorithm using simulations.
Fichier principal
Vignette du fichier
Drift_detection_hal.pdf (284.36 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02552670 , version 1 (23-04-2020)

Identifiants

Citer

Alexandre Reiffers-Masson. Worst-case drift detection of sensor networks: performances and algorithms. SPCOM 2020 : International Conference on Signal Processing and Communications, Jul 2020, Bangalore, India. ⟨10.1109/SPCOM50965.2020.9179620⟩. ⟨hal-02552670⟩
37 Consultations
126 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More