Synchronization of wireless sensors : review of methodologies. Experience feedback of the very precise GPS solution
Résumé
Measuring numerous or various physical parameters all along a structure often
implies that collected data should be compared or correlated by a central process unit
(that ensures the structure assessing or controlling). That typical configuration of a
sensors’ distribution and particularly wireless sensors network, underlines the need of
precise relative or absolute time stamping of each measurement.
Unfortunately, each sensor of a network is an electronic entity whose time-
base varies because of many factors ( its own quartz drift, temperature influence, age,
etc.). Thus, a synchronization mechanism is needed to reach a time stamping that suits
the required precision. The level of precision in sensor synchronization is directly
determined by the SHM need (from few seconds for temperature to micro-seconds for
acoustic wave propagation).
Many topologies could be considered but a review of the appropriate and case-
dependent methodologies is proposed. The simplest case where each sensor directly
measures and transmits data to a supervisor that ensures the final time-stamping is
limited because this supposed a determinist knowledge of the delay due to acquisition
and the RF transmission (moreover, this configuration excludes local signal
processing like those embedded in smart sensors). Two other topologies are analyzed.
First, the topology where a supervisor performs and distributes a synchronization
mechanism through the sensor network. This could be achieved by means of
appropriate network time protocols or by the exchange of specific packets for sensor
time-base reset or synchronization. Another topology consists in an embedded
synchronization mechanism; each sensor manages its own and absolute time-base
Origine : Fichiers produits par l'(les) auteur(s)