pith. sign in

arxiv: 2505.05030 · v4 · pith:NGE5HFGPnew · submitted 2025-05-08 · 📡 eess.SP · math.OC

Autoregressive Stochastic Clock Jitter Compensation in Analog-to-Digital Converters

Daniele Gerosa , Rui Hou , Vimar Bj\"ork , Ulf Gustavsson , Thomas Eriksson This is my paper

Pith reviewed 2026-05-22 16:54 UTC · model grok-4.3

classification 📡 eess.SP math.OC
keywords clock jitteranalog-to-digital converterautoregressive processKalman filterweighted least squarespilot-assisted compensationSINADR improvement
0
0 comments X

The pith

Modeling stochastic clock jitter in ADCs as a first-order autoregressive process enables two pilot-assisted compensation algorithms for baseband signals.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper models the random timing errors in analog-to-digital converters as a stationary first-order autoregressive process. It develops two algorithms that use known pilot signals to estimate and correct these errors in baseband digitization. One solves a sequence of weighted least-squares problems while the other applies a Kalman filter to use the correlation across successive samples. A conditional maximum-likelihood method estimates the autoregressive parameters from the pilots, and the paper analyzes the resulting linearization errors. Simulations indicate the methods raise signal-to-noise-and-distortion ratio by 1 to 15 dB and reduce error vector magnitude by 0.02 to 1.6 dB, with the Kalman smoother performing best when extra temporal data is available.

Core claim

The stochastic discrete-time clock jitter is modeled as a stationary first-order autoregressive process. Two computationally efficient pilot-assisted dejittering algorithms are proposed for baseband signals: one based on solving a sequence of weighted least-squares problems, and another that exploits the correlated jitter structure via a Kalman filter-based routine. A conditional maximum-likelihood estimator for the autoregressive parameters is provided to support near-optimal Kalman performance when parameters vary, along with a mathematical analysis of induced linearization errors and synthetic simulations that report 1-15 dB SINADR gains and 0.02-1.6 dB EVM gains.

What carries the argument

The first-order autoregressive (AR(1)) model of the stochastic clock jitter, which captures its temporal correlation and directly enables both the weighted least-squares sequence and the Kalman filter dejittering routines.

If this is right

  • The weighted least-squares algorithm supplies a direct, pilot-driven correction for jitter effects without requiring full statistical knowledge beyond the model.
  • The Kalman smoother yields higher performance than the least-squares version by incorporating both past and future observations.
  • The conditional maximum-likelihood parameter estimator keeps the Kalman routine near-optimal even when jitter statistics change slowly over time.
  • Signal quality improves measurably, with SINADR gains ranging from 1 dB to 15 dB according to jitter severity and the density of available pilots.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the AR(1) description matches jitter observed in actual high-speed ADCs, the same pilot-assisted routines could be added to existing digital post-processing pipelines to raise effective resolution without new analog hardware.
  • Varying pilot density in the simulations suggests a practical trade-off between compensation accuracy and the fraction of bandwidth spent on training symbols.
  • The linearization-error analysis supplies a template for bounding similar first-order approximations in other timing-recovery or sampling-correction schemes.

Load-bearing premise

The clock jitter follows a stationary first-order autoregressive process whose parameters can be estimated from pilot signals.

What would settle it

Measurements on real ADC hardware with recorded jitter that show no consistent SINADR or EVM improvement when the proposed algorithms are applied compared with uncompensated conversion would falsify the central performance claim.

Figures

Figures reproduced from arXiv: 2505.05030 by Daniele Gerosa, Rui Hou, Thomas Eriksson, Ulf Gustavsson, Vimar Bj\"ork.

Figure 1
Figure 1. Figure 1: Level curves of the term multiplying σ 2 x′ in (12); φ = 0.95 left, φ = 0.999 right, in dB. The double integral in (12) was numerically evaluated via nested integral in MATLAB and its accuracy was double-checked via a separate Monte Carlo where instances of the processes x, ξ etc. were generated and the left-hand side of (12) estimated. subtracted, the pilots removed, and the signal interpolated us￾ing e.g… view at source ↗
Figure 2
Figure 2. Figure 2: ; the focus of this paper is within the green box. Introduce known analog stimulus at transmitter side. Use pilot samples to estimate ξn ∀ n and compensate the distorted signal. Remove sample pilots (if unwanted) and interpolate (e.g. via Gerchberg￾Papoulis) [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Jitter trackers comparison example, Algorithm 1 (left) [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: From top to bottom: σξ/Ts = 5 · 10−3 , 1.5 · 10−2 respectively. NDR is fixed to = −10 dB. Both techniques benefit from the higher pilot density, with (a) σξ/Ts = 5 · 10−3 , φ = 0.999 (b) σξ/Ts = 1.5 · 10−2 , φ = 0.999 [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 4
Figure 4. Figure 4: Varying NDR analysis for two different jitter levels. The white noise power is progressively increased. The “No jitter, [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 7
Figure 7. Figure 7: Algorithm performances vs signal bandwidth. Fixed [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The plot shows a negligible performance loss when [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Power spectral density illustrations at different com [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: EVM performances. On the x axis: pilot-symbol density for Algorithms 1 and 2, and pilot-tone fractional power for [40]. φ = 0.999. Algorithms 1 and 2 display a qualitative behavior consistent with that seen in [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
read the original abstract

This paper addresses the mathematical modeling and compensation of stochastic discrete-time clock jitter in analog-to-digital converters (ADCs). We model the stochastic clock jitter as a first-order autoregressive (AR(1)) process, and we propose two novel, computationally efficient, pilot-assisted dejittering algorithms for baseband signals: one based on solving a sequence of weighted least-squares problems, and another that exploits the correlated jitter structure via a Kalman filter-based routine. We also propose a conditional maximum-likelihood estimator for the autoregressive parameters, enabling near-optimal Kalman-filter performance even when such parameters vary over time. We further provide a mathematical analysis of the induced linearization errors, and we complement the theory with synthetic simulations to evaluate the proposed techniques across different scenarios. The proposed techniques are shown to yield a 1-15 dB improvement in signal-to-noise-and-distortion ratio (SINADR) and 0.02-1.6 dB in symbol error vector magnitude (EVM), depending on impairment severity and pilot density. The Kalman smoother generally provides superior performance by leveraging additional temporal information.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper models stochastic discrete-time clock jitter in ADCs as a stationary first-order autoregressive (AR(1)) process. It proposes two pilot-assisted dejittering algorithms for baseband signals—a sequence of weighted least-squares problems and a Kalman filter routine—along with a conditional maximum-likelihood estimator for the AR coefficients. A mathematical analysis of induced linearization errors is provided, and the methods are evaluated via synthetic Monte-Carlo simulations that report 1–15 dB SINADR gains and 0.02–1.6 dB EVM improvement depending on impairment severity and pilot density, with the Kalman smoother performing best.

Significance. If the AR(1) model and its parameter estimates prove representative of real ADC clock jitter, the proposed algorithms could supply low-complexity, pilot-based compensation techniques that exploit temporal correlation to improve SINADR and EVM in high-speed sampling systems. The explicit derivation of the filters from the AR(1) correlation structure and the accompanying linearization-error analysis constitute clear technical contributions; the availability of synthetic simulation code would further strengthen reproducibility.

major comments (2)
  1. [Simulation Results] Simulation section (Monte-Carlo trials): all reported SINADR/EVM gains and linearization-error results are obtained by sampling jitter directly from the same stationary AR(1) process used to derive the weighted LS and Kalman routines. This setup does not probe performance under model mismatch (e.g., AR(2), non-stationary increments, or measured hardware jitter spectra) and therefore leaves the claimed 1–15 dB improvement range unverified for practical ADCs.
  2. [Mathematical Analysis] § on linearization-error analysis: the error bounds are derived under the assumption that the AR(1) coefficients are known exactly; the propagation of estimation error from the conditional ML estimator into these bounds is not quantified, which is load-bearing for the claim of near-optimal Kalman performance when parameters vary over time.
minor comments (2)
  1. Clarify in the abstract and introduction whether the weighted least-squares formulation reduces to ordinary least-squares when the AR coefficient is zero, and add a brief comparison of computational complexity (flops per sample) between the two proposed algorithms.
  2. Figure captions should explicitly state the pilot density and AR coefficient values used in each curve so that the dependence of the 1–15 dB range on these parameters is immediately visible.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and positive assessment of the technical contributions. We address the two major comments below, proposing targeted revisions to strengthen the manuscript while remaining faithful to the scope of the work.

read point-by-point responses

  1. Referee: [Simulation Results] Simulation section (Monte-Carlo trials): all reported SINADR/EVM gains and linearization-error results are obtained by sampling jitter directly from the same stationary AR(1) process used to derive the weighted LS and Kalman routines. This setup does not probe performance under model mismatch (e.g., AR(2), non-stationary increments, or measured hardware jitter spectra) and therefore leaves the claimed 1–15 dB improvement range unverified for practical ADCs.

    Authors: We agree that the Monte-Carlo results are generated under the exact AR(1) model assumed in the derivations, which is the appropriate setting for validating the proposed weighted least-squares and Kalman-filter algorithms along with the accompanying analysis. This matches standard practice when introducing a new stochastic model and associated estimators. To address concerns about practical applicability, we will revise the simulation section and conclusions to explicitly discuss the modeling assumptions, note the absence of mismatch experiments as a limitation, and outline directions for future validation using AR(2) processes or measured hardware jitter spectra. No new simulation data will be added at this stage, as the current results are intended to demonstrate gains under the stated model. revision: partial

  2. Referee: [Mathematical Analysis] § on linearization-error analysis: the error bounds are derived under the assumption that the AR(1) coefficients are known exactly; the propagation of estimation error from the conditional ML estimator into these bounds is not quantified, which is load-bearing for the claim of near-optimal Kalman performance when parameters vary over time.

    Authors: The linearization-error bounds are derived under known coefficients to obtain closed-form expressions that isolate the effect of the linearization approximation itself. The conditional maximum-likelihood estimator is proven consistent in the manuscript, and the simulation results already show that the Kalman smoother maintains strong performance when parameters are estimated rather than known. We will revise the mathematical analysis section to include a short remark on the impact of parameter estimation error, supported by additional Monte-Carlo trials that compare performance with known versus estimated coefficients. This will better support the claim of near-optimal behavior under time-varying parameters. revision: yes

Circularity Check

0 steps flagged

No circularity: AR(1) modeling is an explicit assumption; algorithms and simulations are derived and evaluated consistently but not tautological.

full rationale

The paper states the AR(1) model as a modeling choice that enables the weighted least-squares and Kalman routines, with parameters estimated via conditional ML from pilots. The mathematical analysis of linearization errors and the performance claims (SINADR/EVM gains) are obtained from derivations and Monte-Carlo trials under the stated model. This is standard model-based evaluation and does not reduce any claimed result to a fitted input or self-citation by construction. No load-bearing self-citations, uniqueness theorems, or ansatzes imported from prior author work are present in the provided text. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that clock jitter follows an AR(1) process whose parameters are estimable from pilots; no free parameters are explicitly fitted beyond the AR coefficients, and no new physical entities are introduced.

free parameters (1)
  • AR(1) coefficients
    Estimated on-line via conditional maximum-likelihood from pilot symbols; treated as time-varying but locally stationary.
axioms (1)
  • domain assumption Stochastic clock jitter is a discrete-time first-order autoregressive process
    Invoked at the opening of the modeling section to justify both estimators.

pith-pipeline@v0.9.0 · 5731 in / 1425 out tokens · 67346 ms · 2026-05-22T16:54:49.310774+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages

  1. [1]

    Aitken,On Least Squares and Linear Combinations of Observations, Proceedings of the Royal Society of Edinburgh, 55, pp

    A. Aitken,On Least Squares and Linear Combinations of Observations, Proceedings of the Royal Society of Edinburgh, 55, pp. 42–48, 1936

    work page 1936

  2. [2]

    Araghi, M

    H. Araghi, M. A. Akhaee, and A. Amini,Joint Compensation of Jitter Noise and Time-Shift Errors in Multichannel Sampling System, IEEE Transactions on Instrumentation and Measurements, vol. 68(10), pp. 3932–3941, 2019

    work page 2019

  3. [3]

    Y . K. Belyaev,Local Properties Of The Sample Functions Of Stationary Gaussian Processes, Theory of Probability & Its Applications, vol. 5(1), pp. 117–123, 1960

    work page 1960

  4. [4]

    Blu, and M

    T. Blu, and M. Unser,Optimal Interpolation of Fractional Brownian Motion Given Its Noisy Samples,2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings, 2006

    work page 2006

  5. [5]

    V . I. Bogachev,Measure Theory, Springer Berlin, 2006

    work page 2006

  6. [6]

    Chang, C.-L

    Y .-S. Chang, C.-L. Lin, W.-S. Wang, C.-C. Lee and C.-Y . Shih,An Analytical Approach for Quantifying Clock Jitter Effects in Continuous- Time Sigma–Delta Modulators, IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 53(9), pp. 1861–1868, 2006

    work page 2006

  7. [7]

    Colone, D

    F. Colone, D. W. O’Hagan, P. Lombardo, and C. J. Baker,A Multistage Processing Algorithm for Disturbance Removal and Target Detection in Passive Bistatic Radar ,IEEE Transactions on Aerospace and Electronic Systems, vol. 45(2), pp. 698–722, 2009

    work page 2009

  8. [8]

    Da Dalt, M

    N. Da Dalt, M. Harteneck, C. Sandner, and A. Wiesbauer,Numerical modeling of PLL jitter and the impact of its non-white spectrum on the SNR of sampled signals, 2001 Southwest Symposium on Mixed-Signal Design (Cat. No.01EX475), pp. 38–44, 2001

    work page 2001

  9. [9]

    V . Divi, G. W. Wornell,Blind Calibration of Timing Skew in Time- Interleaved Analog-to-Digital Converters, IEEE Journal of Selected Topics In Signal Processing, vol. 3(3), pp. 509–522, 2009

    work page 2009

  10. [10]

    Elbornsson,Analysis, Estimation and Compensation of Mismatch Effects in A/D Converters, Ph.D thesis, Department of Electrical En- gineering, Linköping University, 2003

    J. Elbornsson,Analysis, Estimation and Compensation of Mismatch Effects in A/D Converters, Ph.D thesis, Department of Electrical En- gineering, Linköping University, 2003

    work page 2003

  11. [11]

    Eng,Non-Uniform Sampling in Statistical Signal Processing, Ph.D thesis, Department of Electrical Engineering, Linköping University, 2007

    F. Eng,Non-Uniform Sampling in Statistical Signal Processing, Ph.D thesis, Department of Electrical Engineering, Linköping University, 2007

    work page 2007

  12. [12]

    Gävert, M

    B. Gävert, M. Coldrey, and T. Eriksson,Phase noise estimation in OFDM systems, IEEE Transactions on Communications, 2024

    work page 2024

  13. [13]

    R. W. Gerchberg,Super-Resolution through Error Energy Reduction, Optica Acta, vol. 21(9), pp. 709–720, 1974

    work page 1974

  14. [14]

    Gerosa, L

    D. Gerosa, L. Anttila, and T. Eriksson,Compensation of correlated autoregressive clock jitter in arrays of Analog-to-Digital Converters,to appear in the Proceedings of the 59th Asilomar Conference on Signals, Systems, and Computers, 2025

    work page 2025

  15. [15]

    J. A. Grubner,Probability and Random Processes for Electrical and Computer Engineers, Cambridge University Press, 2006

    work page 2006

  16. [16]

    Hernandez, A

    L. Hernandez, A. Wiesbauer, S. Paton, and A. Di Giandomencio, Modelling and optimization of low pass continuous-time sigma delta modulators for clock jitter noise reduction, 2004 IEEE International Symposium on Circuits and Systems (ISCAS), pp. I–1072, 2004

    work page 2004

  17. [17]

    O. G. Ibarra-Manzano, J. A. Andrade-Lucio, M. A. Vazquez Olguin and Y . S. Shmaliy,Kalman Filter for Discrete Processes With Timing Jitter, IEEE Transactions on Signal Processing, vol. 73, pp. 219–229, 2025

    work page 2025

  18. [18]

    Kailath, A

    T. Kailath, A. H. Sayed, and N. Hassibi,Linear Estimation, Prentice Hall, 2000

    work page 2000

  19. [19]

    Kay,Fundamentals of Statistical Signal Processing - Estimation Theory, Pearson, 1993

    S. Kay,Fundamentals of Statistical Signal Processing - Estimation Theory, Pearson, 1993

    work page 1993

  20. [20]

    M. R. Khanzadi, R. Krishnan, D. Kuylenstierna, and T. Eriksson,Os- cillator phase noise and small-scale channel fading in higher frequency bands, 2014 IEEE Globecom Workshops

    work page 2014

  21. [21]

    B. P. Lathi,Modern Digital and Analog Communication Systems (3rd), Oxford University Press, 1998

    work page 1998

  22. [22]

    J. Ma, R. Tao, Y . Li, and X. Kang,Fractional Spectrum Analysis for Nonuniform Sampling in the Presence of Clock Jitter and Timing Offset, IEEE Transactions on Signal Processing, vol. 68, pp. 4148–4162, 2020

    work page 2020

  23. [23]

    Marvasti,Nonuniform sampling: Theory and Practice, Plenum Pub- lishers Co., 2001

    F. Marvasti,Nonuniform sampling: Theory and Practice, Plenum Pub- lishers Co., 2001

    work page 2001

  24. [24]

    M. H. Moghaddam, S. R. Aghdam, and T. Eriksson,An Additive Noise Modeling Technique for Accurate Statistical Study of Residual RF Hardware Impairments, 2019 IEEE International Conference on Communications Workshops (ICC Workshops), Shanghai, China, 2019, pp. 1–5

    work page 2019

  25. [25]

    A. V . Oppenheim, and R. W. Schafer,Discrete-Time Signal Processing, Second Edition, Prentice Hall, 1999

    work page 1999

  26. [26]

    Papoulis,A New Algorithm in Spectral Analysis and Band-Limited Extrapolation,IEEE Transactions on Circuits and Systems, vol

    A. Papoulis,A New Algorithm in Spectral Analysis and Band-Limited Extrapolation,IEEE Transactions on Circuits and Systems, vol. 22, no. 9, pp. 735–742, 1975

    work page 1975

  27. [27]

    Papoulis,Probability, Random V ariables and Stochastic Processes, McGraw-Hill Inc., 1984

    A. Papoulis,Probability, Random V ariables and Stochastic Processes, McGraw-Hill Inc., 1984

    work page 1984

  28. [28]

    Petrovic, W

    B. Petrovic, W. Rave, and G. Fettweis,Phase Noise Suppression in OFDM Using a Kalman Filter, Legacy V odafone chair publication, https://www.vodafone-chair.org/pbls/legacy/d- petrovic/Phase_Noise_Suppression_in_OFDM_using_a_Kalman_Filter.pdf

    work page

  29. [29]

    J. G. Proakis,Digital Communications, McGraw-Hill, 2001

    work page 2001

  30. [30]

    Quarteroni, R

    A. Quarteroni, R. Sacco, and F. Saleri,Numerical Mathematics, Springer, 2007

    work page 2007

  31. [31]

    Nordio, C.-F

    A. Nordio, C.-F. Chiasserini, and E. Viterbo,Signal Reconstruction Errors in Jittered Sampling, IEEE Transactions On Signal Processing, vol. 57(12), pp. 4711–4718, 2009

    work page 2009

  32. [32]

    C. R. Rao, and H. Toutenberg,Linear Models: Least Squares and Alternatives, Springer, 1999

    work page 1999

  33. [33]

    Rutten, L

    R. Rutten, L. J. Breems, and R. H. M. van Veldhoven,Digital jitter- cancellation for narrowband signals, 2008 IEEE International Sympo- sium on Circuits and Systems (ISCAS), pp. 1444–1447, 2008

    work page 2008

  34. [34]

    Salib, M

    A. Salib, M. F. Flanagan, and B. Cardiff,A High-Precision Time Skew Estimation and Correction Technique for Time-Interleaved ADCs, IEEE Transactions on Circuits and Systems-I: Regular Papers, vol. 66(10), pp. 3747–3760, 2019

    work page 2019

  35. [35]

    Shinagawa, Y

    M. Shinagawa, Y . Akazawa, and T. Wakimoto,Jitter analysis of high- speed sampling systems, IEEE Journal of Solid-State Circuits, vol. 25(1), pp. 220–224, 1990

    work page 1990

  36. [36]

    Simon,Optimal State Estimation: Kalman,H ∞, and Nonlinear Approaches, John Wiley & Sons, 2006

    D. Simon,Optimal State Estimation: Kalman,H ∞, and Nonlinear Approaches, John Wiley & Sons, 2006

    work page 2006

  37. [37]

    Syrjälä, and M

    V . Syrjälä, and M. Valkama,Sampling jitter estimation and mitigation in direct RF sub-sampling receiver architecture, 2009 6th International Symposium on Wireless Communication Systems, pp. 323–327, 2006

    work page 2009

  38. [38]

    Sung, and J

    J. Sung, and J. Choi,A New Interpretation of the Time-Interleaved ADC Mismatch Problem: A Tracking-Based Hybrid Calibration Approach, IEEE Signal Processing Letters, vol. 32, pp. 3710–3714, 2025

    work page 2025

  39. [39]

    Testoni, N

    N. Testoni, N. Speciale, A. Ridolfi, and C. Pouzat,Adaptive Wavelet- based signal dejittering, 2007 Ph.D Research in Microelectronics and Electronics Conference, pp. 257–260, 2007

    work page 2007

  40. [40]

    Z. J. Towfic, S.-K. Ting, and A. H. Sayed,Clock Jitter Compensation in High-Rate ADC Circuits, IEEE Transactions on Signal Processing, vol. 60(11), pp. 5738–5753, 2012

    work page 2012

  41. [41]

    Valkama, A

    M. Valkama, A. Springer, and G. Hueber,Digital signal processing for reducing the effects of RF imperfections in radio devices - An overview, Proceedings of 2010 IEEE International Symposium on Circuits and Systems

    work page 2010

  42. [42]

    D. S. Weller, and V . K. Goyal,Bayesian Post-Processing Methods for Jitter Mitigation in Sampling, IEEE Transactions On Signal Processing, vol. 59(5), pp. 2112–2123, 2011

    work page 2011

  43. [43]

    Y . Yang, X. Zhou, and M. Wang,Taylor’s Theorem and Mean V alue Theorem for Random Functions and Random V ariables, https://arxiv.org/abs/2102.10429v2, 2025

    work page arXiv 2025

  44. [44]

    Zanchi, and C

    A. Zanchi, and C. Samori,Analysis and Characterization of the Effects of Clock Jitter in A/D Converters for Subsampling, IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 55(2), pp. 522–534, 2008. APPENDIX Proof of Proposition II.1.This follows from basic properties of random variables. Indeed if var(x′ n) =σ 2 x′, it is well-known

    work page 2008

  45. [45]

    small jitter

    that var(x ′′ n) = 12π2σ2 x′W 2/5, under the assumption that σ2 x = 1andS x(f) = 1/(2W)χ {|f|≤W} (f). The two processesξ 2 n andx ′′ n originate from unrelated sources and can thus be considered independent. Therefore we have var(ξ2 nx′′ n) =var(ξ 2 n)var(x′′ n) +var(ξ 2 n)E[x ′′ n]2 | {z } =0 +var(x′′ n)E[ξ2 n]2 = 36π2σ4 ξ σ2 x′W 2/5; similarly var(ξnx′ ...

    work page