• DocumentCode
    1496238
  • Title

    Stability-Optimized Time Adjustment for a Networked Computer Clock

  • Author

    Yamashita, Takao

  • Author_Institution
    NTT Inf. Sharing Platform Labs., NTT Corp., Musashino, Japan
  • Volume
    59
  • Issue
    10
  • fYear
    2010
  • Firstpage
    1320
  • Lastpage
    1336
  • Abstract
    We propose an optimal time adjustment method from the viewpoint of frequency stability, which is defined as the Allan deviation. When time adjustment is needed for a clock in a networked computer, it is made over a period called a time adjustment period. The proposed method optimizes frequency stability for a given time adjustment period. This method has been evaluated and compared with the adjtime() system call in UNIX systems in terms of frequency stability and duration of time adjustment period needed for achieving particular values of frequency stability. For time intervals from 1 to 1,000 s, the frequency stability achieved by the proposed method was about 0.01-0.5 of that achieved by the adjtime() system call. The evaluation also showed that the duration of a time adjustment period needed for achieving the frequency stability of 1.0 × 10-10 in the proposed method was less than 1/12 (1/6) that of the period in the adjtime() system call when we optimized frequency stability for a 60 (3,600) s time interval under the condition that the duration of the time-adjustment period was 12 h.
  • Keywords
    Unix; clocks; frequency stability; optimisation; synchronisation; Allan deviation; UNIX systems; frequency stability; networked computer clock; stability optimized time adjustment; time adjustment period; Application software; Clocks; Collaboration; Computer networks; Distributed computing; Frequency measurement; Frequency synchronization; Optimization methods; Stability; Time measurement; Allan deviation; Allan variance; Clock synchronization; frequency stability.;
  • fLanguage
    English
  • Journal_Title
    Computers, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9340
  • Type

    jour

  • DOI
    10.1109/TC.2010.106
  • Filename
    5467048