• DocumentCode
    145649
  • Title

    A recursive method for bias estimation in asymmetric packet-based networks

  • Author

    Hajikhani, Maryam ; Kunz, Thomas ; Schwartz, Howard

  • Author_Institution
    Dept. of Syst. & Comput. Eng., Carleton Univ., Ottawa, ON, Canada
  • fYear
    2014
  • fDate
    22-26 Sept. 2014
  • Firstpage
    47
  • Lastpage
    52
  • Abstract
    Clock synchronization in many protocols such as IEEE 1588 is achieved by exchanging timing information between a master and slave node. Packet delay variation (PDV) is a major source of inaccuracy in packet-based synchronization systems. When the expected values of the delays from master to slave and from slave to master are not equal, the synchronization problem can be modeled as a biased estimation problem. In this paper we propose a solution to estimate the delay bias and use this estimate to improve the synchronization accuracy. Our method is easy to implement and is compatible with the current version of the protocol. Moreover, this method allows us to update the slave clock recursively rather than after collecting many samples. The proposed method works well in the presence of frequency offset and does not require any assumption on the filter which is used in the synchronization process.
  • Keywords
    computer networks; protocols; recursive estimation; synchronisation; IEEE 1588; PDV; asymmetric packet-based networks; bias estimation; biased estimation problem; clock synchronization; delay bias; master node; packet delay variation; packet-based synchronization systems; protocols; recursive method; slave clock; slave node; synchronization accuracy; Clocks; Delays; Equations; Estimation; Frequency estimation; Mathematical model; Synchronization; Boot-strap method; IEEE 1588; asymmetric delays; two-way message exchange mechanism;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Precision Clock Synchronization for Measurement, Control, and Communication (ISPCS), 2014 IEEE International Symposium on
  • Conference_Location
    Austin, TX
  • Print_ISBN
    978-1-4799-2698-5
  • Type

    conf

  • DOI
    10.1109/ISPCS.2014.6948530
  • Filename
    6948530