On minimum variance unbiased estimation of clock offset in a two-way message exchange mechanism

Qasim M. Chaudhari, Erchin Serpedin, Khalid Qaraqe

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

For many applications, distributed networks require the local clocks of the constituent nodes to run close to an agreed upon notion of time. Most of the widely used clock synchronization algorithms in such systems employ the sender-receiver protocol based on a two-way timing message exchange paradigm. Maximum likelihood estimator (MLE) of the clock offset based on the timing message exchanges between two clocks was derived in D. R. Jeske, On maximum likelihood estimation of clock offset[IEEE Trans. Commun., vol. 53, pp. 53-54, Jan. 2005], when the fixed delays are symmetric and the variable delays in each direction assume an exponential distribution with an unknown mean. Herein, the best linear unbiased estimate using order statistics (BLUE-OS) of the clock offset between two nodes is derived assuming both symmetric and asymmetric exponential network delays, respectively. The Rao-Blackwell-Lehmann- Scheffé theorem is then exploited to obtain the minimum variance unbiased estimate (MVUE) for the clock offset which it is shown to coincide with the BLUE-OS. In addition, it is found that the MVUE of the clock offset in the presence of symmetric network delays also coincides with the MLE. Finally, in the presence of asymmetric network delays, although the MLE is biased, it is shown to achieve lesser mean-square error (MSE) than the MVUE in the region around the point where the bidirectional network link delays are symmetric and hence its merit as the most versatile estimator is fairly justified.

Original languageEnglish
Article number2046233
Pages (from-to)2893-2904
Number of pages12
JournalIEEE Transactions on Information Theory
Volume56
Issue number6
DOIs
Publication statusPublished - Jun 2010
Externally publishedYes

Fingerprint

Clocks
Maximum likelihood
statistics
Statistics
recipient
Maximum likelihood estimation
paradigm
Mean square error
Synchronization

Keywords

  • Clock
  • Estimation
  • Signal processing
  • Synchronization
  • Wireless sensor networks

ASJC Scopus subject areas

  • Information Systems
  • Computer Science Applications
  • Library and Information Sciences

Cite this

On minimum variance unbiased estimation of clock offset in a two-way message exchange mechanism. / Chaudhari, Qasim M.; Serpedin, Erchin; Qaraqe, Khalid.

In: IEEE Transactions on Information Theory, Vol. 56, No. 6, 2046233, 06.2010, p. 2893-2904.

Research output: Contribution to journalArticle

@article{ef3a833c0f6a463bb80edd04925c7c63,
title = "On minimum variance unbiased estimation of clock offset in a two-way message exchange mechanism",
abstract = "For many applications, distributed networks require the local clocks of the constituent nodes to run close to an agreed upon notion of time. Most of the widely used clock synchronization algorithms in such systems employ the sender-receiver protocol based on a two-way timing message exchange paradigm. Maximum likelihood estimator (MLE) of the clock offset based on the timing message exchanges between two clocks was derived in D. R. Jeske, On maximum likelihood estimation of clock offset[IEEE Trans. Commun., vol. 53, pp. 53-54, Jan. 2005], when the fixed delays are symmetric and the variable delays in each direction assume an exponential distribution with an unknown mean. Herein, the best linear unbiased estimate using order statistics (BLUE-OS) of the clock offset between two nodes is derived assuming both symmetric and asymmetric exponential network delays, respectively. The Rao-Blackwell-Lehmann- Scheff{\'e} theorem is then exploited to obtain the minimum variance unbiased estimate (MVUE) for the clock offset which it is shown to coincide with the BLUE-OS. In addition, it is found that the MVUE of the clock offset in the presence of symmetric network delays also coincides with the MLE. Finally, in the presence of asymmetric network delays, although the MLE is biased, it is shown to achieve lesser mean-square error (MSE) than the MVUE in the region around the point where the bidirectional network link delays are symmetric and hence its merit as the most versatile estimator is fairly justified.",
keywords = "Clock, Estimation, Signal processing, Synchronization, Wireless sensor networks",
author = "Chaudhari, {Qasim M.} and Erchin Serpedin and Khalid Qaraqe",
year = "2010",
month = "6",
doi = "10.1109/TIT.2010.2046233",
language = "English",
volume = "56",
pages = "2893--2904",
journal = "IEEE Transactions on Information Theory",
issn = "0018-9448",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "6",

}

TY - JOUR

T1 - On minimum variance unbiased estimation of clock offset in a two-way message exchange mechanism

AU - Chaudhari, Qasim M.

AU - Serpedin, Erchin

AU - Qaraqe, Khalid

PY - 2010/6

Y1 - 2010/6

N2 - For many applications, distributed networks require the local clocks of the constituent nodes to run close to an agreed upon notion of time. Most of the widely used clock synchronization algorithms in such systems employ the sender-receiver protocol based on a two-way timing message exchange paradigm. Maximum likelihood estimator (MLE) of the clock offset based on the timing message exchanges between two clocks was derived in D. R. Jeske, On maximum likelihood estimation of clock offset[IEEE Trans. Commun., vol. 53, pp. 53-54, Jan. 2005], when the fixed delays are symmetric and the variable delays in each direction assume an exponential distribution with an unknown mean. Herein, the best linear unbiased estimate using order statistics (BLUE-OS) of the clock offset between two nodes is derived assuming both symmetric and asymmetric exponential network delays, respectively. The Rao-Blackwell-Lehmann- Scheffé theorem is then exploited to obtain the minimum variance unbiased estimate (MVUE) for the clock offset which it is shown to coincide with the BLUE-OS. In addition, it is found that the MVUE of the clock offset in the presence of symmetric network delays also coincides with the MLE. Finally, in the presence of asymmetric network delays, although the MLE is biased, it is shown to achieve lesser mean-square error (MSE) than the MVUE in the region around the point where the bidirectional network link delays are symmetric and hence its merit as the most versatile estimator is fairly justified.

AB - For many applications, distributed networks require the local clocks of the constituent nodes to run close to an agreed upon notion of time. Most of the widely used clock synchronization algorithms in such systems employ the sender-receiver protocol based on a two-way timing message exchange paradigm. Maximum likelihood estimator (MLE) of the clock offset based on the timing message exchanges between two clocks was derived in D. R. Jeske, On maximum likelihood estimation of clock offset[IEEE Trans. Commun., vol. 53, pp. 53-54, Jan. 2005], when the fixed delays are symmetric and the variable delays in each direction assume an exponential distribution with an unknown mean. Herein, the best linear unbiased estimate using order statistics (BLUE-OS) of the clock offset between two nodes is derived assuming both symmetric and asymmetric exponential network delays, respectively. The Rao-Blackwell-Lehmann- Scheffé theorem is then exploited to obtain the minimum variance unbiased estimate (MVUE) for the clock offset which it is shown to coincide with the BLUE-OS. In addition, it is found that the MVUE of the clock offset in the presence of symmetric network delays also coincides with the MLE. Finally, in the presence of asymmetric network delays, although the MLE is biased, it is shown to achieve lesser mean-square error (MSE) than the MVUE in the region around the point where the bidirectional network link delays are symmetric and hence its merit as the most versatile estimator is fairly justified.

KW - Clock

KW - Estimation

KW - Signal processing

KW - Synchronization

KW - Wireless sensor networks

UR - http://www.scopus.com/inward/record.url?scp=77957587744&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77957587744&partnerID=8YFLogxK

U2 - 10.1109/TIT.2010.2046233

DO - 10.1109/TIT.2010.2046233

M3 - Article

AN - SCOPUS:77957587744

VL - 56

SP - 2893

EP - 2904

JO - IEEE Transactions on Information Theory

JF - IEEE Transactions on Information Theory

SN - 0018-9448

IS - 6

M1 - 2046233

ER -