Timestamp-Free Clock Syntonization for IoT Using Carrier Frequency Offset

Baofeng Zhou, Fujuan Guo, Mehmet C. Vuran

Research output: Contribution to journalArticlepeer-review


System-level timing fluctuations caused by unstable low-cost clocks and end-to-end communication delays are the main sources of uncertainties in existing synchronization mechanisms that rely on timestamp exchanges. This paper introduces a timestamp-free clock syntonization approach, carrier frequency offset (CFO)-assisted syntonization (CFOSynt), to estimate the clock skew between a pair of nodes by utilizing carrier frequency offset. To enable CFOSynt, we leverage the fact that RF oscillators in the radio can be used as the reference to calibrate the system clock oscillators, and the pairwise RF clock information is carried in the transmission carrier frequency. By incorporating CFO and capturing the clock frequency relationship in system clock skew estimation, CFOSynt can eliminate the need for timestamping and the impact of delay uncertainties. To validate the design, CFOSynt is implemented on two common off-the-shelf (COTS) IoT platforms with access to the CFO estimation from the radio chip. Extensive experiments are conducted to evaluate CFOSynt, and CFOSynt can estimate the clock skew of 32 kHz low-cost electronic oscillators with a mean error of $-2.46$-2.46 Hz. In comparison with timestamp-based approaches, CFOSynt achieves up to 70-90 percent improvement in skew estimation error and shows significant reliability when low-cost oscillators are used.

Original languageEnglish (US)
Pages (from-to)712-727
Number of pages16
JournalIEEE Transactions on Mobile Computing
Issue number2
StatePublished - Feb 1 2022


  • carrier frequency offset
  • clock synchronization
  • Internet of Things
  • timestamp-free syntonization

ASJC Scopus subject areas

  • Software
  • Computer Networks and Communications
  • Electrical and Electronic Engineering


Dive into the research topics of 'Timestamp-Free Clock Syntonization for IoT Using Carrier Frequency Offset'. Together they form a unique fingerprint.

Cite this