Joint Orientation and Position Estimation from Differential RSS Measurements at On-Body Nodes - Université de Rennes Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Joint Orientation and Position Estimation from Differential RSS Measurements at On-Body Nodes

Résumé

In the specific context of wearable wireless networks, this work aims at improving the accuracy and the robustness of radiolocation solutions based on standard narrow-band radio technologies for user-centric mobile applications. The proposed solution takes benefits from off-body or body-to-body radio links, with respect to fixed elements of infrastructure or to other mobile equipped subjects, respectively. The main idea is to infer relative angular information between the carrying body's heading and the median direction of arrival of received signals. For this sake, we rely on differential received power measurements issued at judiciously placed on-body nodes. Light calibration procedures (e.g., based on the full-scale dynamics of observed measurements) and joint absolute position/orientation estimation algorithms then enable to cover both individual/non-cooperative and collective/cooperative navigation needs. The performance is assessed by means of field experiments with IEEE 802.15.4-compliant integrated devices operating at 2.4 GHz and an optical motion capture system delivering the ground truth reference. On this occasion, our proposal is shown to be resilient against mobile Non-line of Sight (NLoS) situations (e.g., in crowded environments).
Fichier non déposé

Dates et versions

hal-01479175 , version 1 (28-02-2017)

Identifiants

  • HAL Id : hal-01479175 , version 1

Citer

B. Denis, Bernard Uguen, F. Mani, Raffaele d'Errico, N. Amiot. Joint Orientation and Position Estimation from Differential RSS Measurements at On-Body Nodes. Ieee 27th Annual International Symposium On Personal, Indoor, and Mobile Radio Communications (pimrc), Sep 2016, Valencia, Spain. pp.1321--1326. ⟨hal-01479175⟩
133 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More