{
  "id": "allan-deviation",
  "title": "Allan Deviation",
  "lang": "en",
  "source_lang": "en",
  "summary": "Allan deviation measures signal stability across averaging times by comparing successive averages, especially in clocks, oscillators, and precision sensors.",
  "category": "science",
  "subcategory": "physics",
  "type": "concept",
  "aliases": [
    "Allan standard deviation",
    "ADEV"
  ],
  "generated_at": "2026-11-05",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/allan-deviation/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/allan-deviation/",
      "json": "https://aiwiki.page/zh/allan-deviation.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Allan deviation is a statistical measure of how the stability of a measured signal changes with averaging time. It is the square root of the Allan variance, which is defined using differences between successive, equally spaced averages rather than deviations from a single overall mean. Developed for characterizing frequency stability in clocks and oscillators, it is widely used in metrology, precision sensing, and navigation. Its central feature is that it distinguishes fluctuations occurring on different timescales and remains useful for several common noise processes for which ordinary variance does not approach a finite, record-length-independent value."
    },
    {
      "heading": "Definition",
      "text": "For an oscillator with instantaneous frequency \\(\\nu(t)\\) and nominal frequency \\(\\nu_0\\), the fractional frequency deviation is\n\n\\[y(t)=\\frac{\\nu(t)-\\nu_0}{\\nu_0}.\\]\n\nThe average fractional frequency over an interval of duration \\(\\tau\\) is\n\n\\[\\bar y_k(\\tau) = \\frac{1}{\\tau} \\int_{t_k}^{t_k+\\tau}y(t)\\,dt,\\]\n\nwhere consecutive intervals are contiguous: \\(t_{k+1}=t_k+\\tau\\).\n\nThe Allan variance is\n\n\\[\\sigma_y^2(\\tau) = \\frac{1}{2} \\mathbb{E} \\left[ \\left( \\bar y_{k+1}(\\tau)-\\bar y_k(\\tau) \\right)^2 \\right],\\]\n\nwhere \\(\\mathbb{E}\\) denotes the expected value. The Allan deviation is\n\n\\[\\sigma_y(\\tau)=\\sqrt{\\sigma_y^2(\\tau)}.\\]\n\nThe averaging time \\(\\tau\\) is an essential part of the result. A quoted Allan deviation without its associated averaging time does not fully specify a stability measurement.\n\nWhen calculated from fractional frequency, Allan deviation is dimensionless. For example, \\(\\sigma_y(1\\,\\mathrm{s})=10^{-12}\\) describes fractional frequency fluctuations on a one-second averaging timescale. Multiplying by \\(\\nu_0\\) expresses the corresponding fluctuation scale in frequency units.\n\nThe same construction can be applied to other measured quantities, such as voltage, angular rate, or acceleration. The resulting deviation then has the units of that quantity, and its interpretation depends on the signal being analyzed."
    },
    {
      "heading": "Why successive differences are used",
      "text": "Ordinary variance measures dispersion around a mean. It is effective for many stationary processes, but oscillator signals often contain substantial low-frequency fluctuations. For some such noise models, the sample variance depends strongly on the duration of the record rather than converging to a fixed value.\n\nAllan variance instead compares neighboring interval averages. This differencing suppresses slowly varying components and removes a constant frequency offset exactly. A clock can therefore have low Allan deviation while running at a consistently incorrect frequency: stability is not the same as accuracy.\n\nFor independent successive averages with equal variance,\n\n\\[\\frac{1}{2} \\mathbb{E} \\left[ (\\bar y_{k+1}-\\bar y_k)^2 \\right] = \\operatorname{Var}(\\bar y_k).\\]\n\nThis explains the factor of one half. More generally, when the averaged sequence is stationary,\n\n\\[\\sigma_y^2(\\tau) = \\operatorname{Var}(\\bar y_k) - \\operatorname{Cov}(\\bar y_{k+1},\\bar y_k).\\]\n\nAllan deviation therefore depends on both the dispersion of interval averages and their correlation."
    },
    {
      "heading": "Estimation from measurements",
      "text": "Non-overlapping estimates\n\nGiven \\(M\\) consecutive, non-overlapping averages of duration \\(\\tau\\), the conventional estimator is\n\n\\[\\widehat{\\sigma}_y^2(\\tau) = \\frac{1}{2(M-1)} \\sum_{k=1}^{M-1} \\left( \\bar y_{k+1}-\\bar y_k \\right)^2.\\]\n\nTo evaluate several averaging times, regularly sampled measurements are grouped into progressively longer intervals. If the basic sampling interval is \\(\\tau_0\\), the averaging times are commonly chosen as \\(\\tau=m\\tau_0\\), where \\(m\\) is a positive integer.\n\nAt long averaging times, fewer independent comparisons remain. The resulting estimates have greater statistical uncertainty.\n\nTime-error representation\n\nClock measurements are often recorded as time error \\(x(t)\\), measured in seconds relative to a reference. Under the usual convention,\n\n\\[y(t)=\\frac{dx(t)}{dt},\\]\n\nso the average fractional frequency becomes\n\n\\[\\bar y_k(\\tau) = \\frac{x(t_k+\\tau)-x(t_k)}{\\tau}.\\]\n\nConsequently,\n\n\\[\\sigma_y^2(\\tau) = \\frac{1}{2\\tau^2} \\mathbb{E} \\left[ \\left( x(t+2\\tau)-2x(t+\\tau)+x(t) \\right)^2 \\right].\\]\n\nAllan variance is thus proportional to the mean square of a second difference of time error. This is particularly useful in clock-comparison systems, where phase or time differences are measured directly.\n\nOverlapping estimates\n\nThe overlapping Allan variance uses all available adjacent pairs of averaging windows, shifting their starting point by one basic sampling interval rather than by an entire averaging time.\n\nFor \\(N\\) time-error samples \\(x_0,\\ldots,x_{N-1}\\), spaced by \\(\\tau_0\\), an overlapping estimator at \\(\\tau=m\\tau_0\\) is\n\n\\[\\widehat{\\sigma}_{y,\\mathrm{ov}}^2(\\tau) = \\frac{1}{2\\tau^2(N-2m)} \\sum_{i=0}^{N-2m-1} \\left( x_{i+2m}-2x_{i+m}+x_i \\right)^2,\\]\n\nprovided \\(N>2m\\).\n\nOverlapping generally improves statistical precision by using more of the available record. The terms in the sum are correlated, however, so their number is not the number of independent observations."
    },
    {
      "heading": "Reading an Allan-deviation plot",
      "text": "Allan deviation is usually plotted against averaging time on logarithmic axes. Different noise processes produce characteristic slopes over ranges where a single process dominates.\n\nFor conventional oscillator-noise models:\n\nDominant process | Approximate dependence of Allan deviation\nWhite phase modulation | \\(\\sigma_y(\\tau)\\propto\\tau^{-1}\\), for fixed measurement bandwidth in the usual asymptotic regime\nFlicker phase modulation | Approximately \\(\\tau^{-1}\\), with bandwidth-dependent logarithmic corrections\nWhite frequency modulation | \\(\\sigma_y(\\tau)\\propto\\tau^{-1/2}\\)\nFlicker frequency modulation | \\(\\sigma_y(\\tau)\\propto\\tau^0\\)\nRandom-walk frequency modulation | \\(\\sigma_y(\\tau)\\propto\\tau^{+1/2}\\)\nLinear frequency drift | \\(\\sigma_y(\\tau)\\propto\\tau\\)\n\n“White” denotes a flat relevant power spectral density, while flicker noise has a spectrum proportional to inverse frequency. Random-walk frequency noise is related to Brownian motion in the frequency variable.\n\nA falling curve indicates that longer averaging reduces the measured fluctuations over that range. A flat region is often associated with flicker frequency noise. A rising curve can reflect random-walk behavior, drift, environmental disturbances, or other slow changes.\n\nThese signatures are not unique diagnoses. Multiple noise sources can coexist, and periodic disturbances can produce humps or oscillatory features. In particular, ordinary Allan deviation does not cleanly distinguish white from flicker phase modulation merely by the leading slope.\n\nFor deterministic linear fractional-frequency drift,\n\n\\[y(t)=y_0+Dt,\\]\n\nthe contribution is\n\n\\[\\sigma_y(\\tau)=\\frac{|D|\\tau}{\\sqrt{2}}.\\]\n\nRemoving fitted drift changes the statistic being reported and should therefore be stated explicitly."
    },
    {
      "heading": "Relationship to spectral analysis",
      "text": "For a fractional-frequency process described by a one-sided power spectral density \\(S_y(f)\\), Allan variance can be written, when the integral converges, as\n\n\\[\\sigma_y^2(\\tau) = 2\\int_0^\\infty S_y(f) \\frac{\\sin^4(\\pi f\\tau)} {(\\pi f\\tau)^2}\\,df.\\]\n\nThis expression connects time-domain stability measurements to frequency-domain analysis. The weighting function combines interval averaging with differencing between adjacent intervals.\n\nAt low Fourier frequencies, the weighting tends to zero in proportion to \\(f^2\\), suppressing slowly varying components. This explains why Allan variance can remain finite for several low-frequency noise models whose ordinary variance diverges.\n\nThe high-frequency response also matters. For phase-noise processes, results can depend strongly on measurement bandwidth. The averaging time alone therefore does not describe the complete measurement system; filtering and sampling characteristics may also need to be specified."
    },
    {
      "heading": "Statistical interpretation and limitations",
      "text": "Stability is not measurement uncertainty\n\nAllan deviation is a measure of fluctuation, not a complete statement of measurement uncertainty. It does not, by itself, account for systematic offsets, calibration errors, reference-frequency errors, or all environmental influences. Nor is it automatically the uncertainty of the average over an entire measurement campaign.\n\nTranslating a stability curve into uncertainty for a particular measurement requires assumptions about the noise processes and the measurement procedure.\n\nFinite records and confidence intervals\n\nAn Allan-deviation estimate is itself uncertain. Its confidence interval depends on record length, averaging time, estimator type, and noise behavior.\n\nFor common Gaussian noise models, intervals for Allan variance are often approximated using a chi-squared distribution with an effective number of degrees of freedom. This number can be much smaller than the number of squared differences, particularly for overlapping estimates or strongly correlated noise.\n\nThe longest-\\(\\tau\\) points are consequently often the least certain. A sharp minimum near the end of a short record need not identify a reproducible optimum averaging time.\n\nSampling and instrument effects\n\nThe standard formulas assume regular sampling and contiguous measurement intervals. Dead time between measurements changes the response to noise and may introduce bias. Missing samples, outliers, frequency steps, and changes in operating conditions can also alter the curve.\n\nFrequency counters differ in their internal weighting and averaging. Data produced by one counter architecture may not represent the rectangular interval averages assumed in a conventional Allan-variance calculation.\n\nA comparison between two clocks measures their combined fluctuations. If the clocks are independent, their Allan variances add; the comparison does not directly yield the instability of either clock alone."
    },
    {
      "heading": "Related stability measures",
      "text": "Several related statistics address limitations or emphasize different features:\n\n- Modified Allan deviation adds an averaging operation that helps distinguish white phase modulation from flicker phase modulation.\n- Time deviation expresses a related measure in units of time. Under the conventional definition,\n\n\\[\\operatorname{TDEV}(\\tau) = \\frac{\\tau}{\\sqrt{3}}\\, \\operatorname{MDEV}(\\tau).\\]\n- Hadamard variance uses second differences of averaged frequency, making it insensitive to linear frequency drift.\n- Total variance and total deviation use extended-data constructions to improve statistical performance, particularly at longer averaging times.\n\nThese quantities are related but not interchangeable. Their different weighting functions produce different responses to noise and drift."
    },
    {
      "heading": "History and applications",
      "text": "The measure is named after David W. Allan, whose 1966 paper on the statistics of atomic frequency standards established its foundational treatment. It became a standard tool in time and frequency metrology as increasingly stable oscillators required methods suited to correlated and low-frequency noise.\n\nIn atomic clocks and oscillator development, Allan deviation characterizes short- and long-term frequency stability. Clock comparisons, satellite-navigation systems, and time-transfer experiments use it to assess timing references and measurement links.\n\nIn precision interferometry and spectroscopy, it helps characterize the stability of lasers, frequency references, and measured signals. In inertial sensors, including gyroscopes and accelerometers, analogous Allan-deviation plots separate noise behavior across timescales. Sensor-specific parameters inferred from these plots depend on the noise model and the conventions used to define those parameters."
    }
  ],
  "links": [
    "metrology",
    "expected-value",
    "second",
    "variance",
    "brownian-motion",
    "measurement-uncertainty",
    "calibration",
    "confidence-interval",
    "atomic-clock",
    "interferometry",
    "spectroscopy"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}