Microsecond-Accuracy Time Synchronization Using the IEEE 802.15.4 TSCH Protocol

Abstract : Time-Slotted Channel Hopping from the IEEE 802.15.4-2015 standard requires that network nodes are tightly time-synchronized. Existing implementations of TSCH on embedded hardware are characterized by tens-of-microseconds large synchronization errors; higher synchronization accuracy would enable reduction of idle listening time on receivers, in this way decreasing the energy required to run TSCH. For some applications, it would also allow to replace dedicated time synchronization mechanisms with TSCH. We show that time synchronization errors in the existing TSCH implementations on embedded hardware are caused primarily by imprecise clock drift estimations, rather than by real unpredictable drift variance. By estimating clock drift more precisely and by applying adaptive time compensation on each node in the network, we achieve microsecond accuracy time synchronization on point-to-point links and a <2 microsecond end-to-end error in a 7-node line topology. Our solution is implemented in the Contiki operating system and tested on Texas Instruments CC2650-based nodes, equipped with common off-the-shelf hardware clock sources (20 ppm drift). Our implementation uses only standard TSCH control messages and is able to keep radio duty cycle below 1%.
Document type :
Conference papers
Complete list of metadatas

Cited literature [8 references]  Display  Hide  Download

https://hal.inria.fr/hal-01351144
Contributor : Simon Duquennoy <>
Submitted on : Thursday, September 22, 2016 - 1:03:28 PM
Last modification on : Monday, September 26, 2016 - 1:21:27 PM

File

tsch-cc2650.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01351144, version 1

Collections

Citation

Atis Elsts, Simon Duquennoy, Xenofon Fafoutis, George Oikonomou, Robert Piechocki, et al.. Microsecond-Accuracy Time Synchronization Using the IEEE 802.15.4 TSCH Protocol. IEEE SenseApp 2016 - Eleventh IEEE International Workshop on Practical Issues in Building Sensor Network Applications, Nov 2016, Dubai, United Arab Emirates. ⟨hal-01351144⟩

Share

Metrics

Record views

377

Files downloads

1219