• DocumentCode
    819627
  • Title

    Ratio indicator characterization for measuring the precision of an estimate obtained by processing sampled data

  • Author

    Hao, Wei-Da ; Jenq, Yih-Chyun ; Wang, Shiang-Yu

  • Author_Institution
    Dept. of Electr. Eng. & Comput. Sci., Texas A&M Univ.-Kingsville, Kingsville, TX, USA
  • Volume
    54
  • Issue
    3
  • fYear
    2005
  • fDate
    6/1/2005 12:00:00 AM
  • Firstpage
    1156
  • Lastpage
    1165
  • Abstract
    This paper aims to characterize the ratio indicator proposed in reference [1]. Ratio indicator enables the stochastic up and down method to have the ability to show users the precision of the estimation, so that waveform reconstruction with jittered data using the said method can be conducted in a reliable way. After studying in detail the stochastic stationary model, we initiate 15 random walks of decreasing threshold intervals and calculate the ninety-eighth percentiles of the corresponding stationary distributions. Analysis on the computing result enables us to find the characterization of ratio indicator. Our target is to advance the usage of ratio indicator. What is proved in the paper is that ratio indicator can be applied more rigorously to show the error bound of the estimation of the median.
  • Keywords
    estimation theory; jitter; random processes; signal reconstruction; signal sampling; stochastic processes; waveform analysis; estimate precision measurement; jittered data; random walks; ratio indicator characterization; stochastic stationary model; stochastic up and down method; waveform reconstruction; Circuit noise; Extraterrestrial measurements; Noise level; Sampling methods; Signal reconstruction; Signal sampling; Stochastic processes; Stochastic resonance; Time measurement; Timing jitter; Markov process; signal reconstruction; signal sampling; stochastic approximation; timing jitter; waveform analysis;
  • fLanguage
    English
  • Journal_Title
    Instrumentation and Measurement, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9456
  • Type

    jour

  • DOI
    10.1109/TIM.2005.847185
  • Filename
    1433189