pith. sign in

arxiv: 2604.11601 · v1 · submitted 2026-04-13 · 📡 eess.SP

The Memory-Enhanced Gaussian Noise (MEGN) Model for Fiber-Optic Channels

Kaiquan Wu , Gabriele Liga , Marco Secondini , Stella Civelli , Hussam Batshon , Greg Raybon , Xi Chen , Alex Alvarado This is my paper

Pith reviewed 2026-05-10 15:06 UTC · model grok-4.3

classification 📡 eess.SP
keywords fiber-optic channelsnonlinear interferenceGaussian noise modelsymbol energy correlationscoded modulationcoherent transmissionmemory-enhanced model
0
0 comments X

The pith

The MEGN model extends the EGN model to estimate nonlinear interference power when symbols exhibit energy correlations over time.

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

The paper derives a mathematical extension to the enhanced Gaussian noise model that incorporates memory effects arising from correlations in the energy of transmitted symbols. This extension matters because recent coded modulation techniques deliberately introduce such correlations to reduce nonlinear interference, which violates the independent and identically distributed symbol assumption of the original model. Validation through both numerical simulations and transmission experiments shows that the new model keeps normalized average NLI power estimation errors below 5 percent across a wide range of symbol rates and distances. The resulting framework supplies a way to analyze and optimize optical transmission systems that employ these temporally correlated modulation schemes.

Core claim

This work presents a rigorous mathematical derivation of the Memory-Enhanced Gaussian Noise (MEGN) model, a memory extension of the EGN model that explicitly accounts for symbol energy correlations in the calculation of accumulated nonlinear interference power for coherent fiber-optic transmission systems.

What carries the argument

The MEGN model, which extends the EGN model's time-invariant signal statistics to include the effects of temporal symbol energy correlations on nonlinear interference.

If this is right

  • Normalized average NLI power estimations remain accurate with less than 5 percent error for a wide range of symbol rates and transmission distances.
  • The model supplies a theoretical framework for analyzing optical transmission systems that use temporally correlated modulation schemes.
  • The framework supports optimization of systems that introduce controlled symbol correlations to mitigate nonlinear interference.
  • The approach applies to both numerical simulations and real transmission experiments.

Where Pith is reading between the lines

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

  • Link design tools could incorporate the MEGN model to set more precise power budgets when using advanced modulation formats.
  • The model could be extended to other forms of temporal correlation if energy proves not to be the dominant factor.
  • Designers might use the framework to test new correlation patterns that further reduce NLI while preserving data rate.

Load-bearing premise

Symbol energy correlations exert the most significant influence on nonlinear interference power among possible temporal correlations.

What would settle it

An experiment or simulation with temporally correlated symbols in which the MEGN model's predicted normalized average NLI power deviates by more than 5 percent from measured values would falsify the accuracy claim.

Figures

Figures reproduced from arXiv: 2604.11601 by Alex Alvarado, Gabriele Liga, Greg Raybon, Hussam Batshon, Kaiquan Wu, Marco Secondini, Stella Civelli, Xi Chen.

Figure 1
Figure 1. Figure 1: (a): Hierarchy of GN-type NLI power models according to the symbol-energy statistics they capture. The proposed [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Organization of the paper structure of the analytical derivations in this work. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: An illustration of the relationship between symbol [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: An illustration of the 1-D, 2-D, and 4-D mapping [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Illustration of the symbol-energy covariances in ( [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Illustration of the MEGN channel functions for SPT in ( [PITH_FULL_IMAGE:figures/full_fig_p012_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The contributions due to SPT, XP, and XPT-type energy [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: The MEGN model accuracy as the memory M in￾creases for N = 400 with 4D mapping, at Rs = 32 GBd. In particular, [PITH_FULL_IMAGE:figures/full_fig_p013_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: (a): NLI PSD reduction GC (f) in (3) with shaping blocklength n = 100, 500 and 1D/2D/4D mappings for transmission at Rs = 32 GBd and Ns = 10 spans. (b): the optimal effective SNR (left axis) and the NLI power coefficient η (right axis) vs. the CCDM shaping blocklength n after transmission of 10 spans, for transmission at Rs = 32 GBd and Ns = 10 spans. The NLI power efficiency obtained from the simulation a… view at source ↗
Figure 10
Figure 10. Figure 10: The comparison of estimate errors ∆η for the EGN and MEGN for various shaping blocklength and 1D (the first row), 2D (the second row) and 4D (the third row) mappings. The first two columns present the accuracy for various number of spans Ns at symbol rate Rs = 32 GBd, while the last two columns present various symbol rates Rs at at Ns = 10 spans [PITH_FULL_IMAGE:figures/full_fig_p015_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: The effective SNR vs. the CCDM shaping blocklength [PITH_FULL_IMAGE:figures/full_fig_p016_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: The division of 2D region for covariance function [PITH_FULL_IMAGE:figures/full_fig_p018_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: An example of the divisions of 2D region for [PITH_FULL_IMAGE:figures/full_fig_p019_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: The support region in D (1) for covariance function K¯ (τ, τ ′ ). The shaded areas are nonzero. can be partitioned as X D K¯ 3 (τ, τ ′ ) = X C1 E h |a| 2 i K¯ 1 (τ ) +X C2 E h |a| 2 i K¯ 1 (τ ′ ) + X C3 E h |a| 2 i K¯ 1 (τ ′ − τ ) +X O h K¯ 3 (τ, τ ′ ) − E h |a| 2 i [PITH_FULL_IMAGE:figures/full_fig_p019_14.png] view at source ↗
read the original abstract

The enhanced Gaussian noise (EGN) model is widely used for estimating the nonlinear interference (NLI) power accumulated in coherent fiber-optic transmission systems. Given a fixed fiber link, under the assumption that transmitted symbols are independently and identically distributed (i.i.d.), the EGN model establishes that the NLI power depends on time-invariant signal statistics, i.e., the second-, fourth-, and sixth-order moments of the symbols, which are determined by the modulation format and its probability distribution. However, recent advances in coded modulation have sought to mitigate NLI by introducing controlled temporal correlations among transmitted symbols, thereby violating the i.i.d. assumption underlying the EGN model. Among these correlations, symbol energy correlations are believed to exert the most significant influence on NLI. This work presents a rigorous mathematical derivation of a memory extension of the EGN model that explicitly accounts for symbol energy correlations, referred to as the MEGN model. The proposed MEGN model is validated through both numerical simulations and transmission experiments. Normalized average NLI power estimations with less than 5% errors across a wide range of symbol rates and transmission distances are reported. The model also provides a theoretical framework for analyzing and optimizing optical transmission systems employing temporally correlated modulation schemes.

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 / 1 minor

Summary. The paper derives the Memory-Enhanced Gaussian Noise (MEGN) model as an extension of the EGN model to account for second-order symbol energy correlations in non-i.i.d. transmitted symbols, while retaining the Gaussian-noise approximation for NLI power estimation in fiber-optic channels. It claims rigorous mathematical derivation and reports validation via numerical simulations and transmission experiments with normalized average NLI power errors below 5% across wide ranges of symbol rates and distances, positioning MEGN as a framework for analyzing and optimizing temporally correlated modulation schemes.

Significance. If the central premise holds, the MEGN model supplies a concrete, parameter-free extension of the widely used EGN framework that enables quantitative analysis of NLI under controlled temporal correlations, directly supporting optimization of coded-modulation formats in coherent optical systems.

major comments (2)
  1. [Abstract] Abstract: the claim that MEGN provides a 'theoretical framework for analyzing and optimizing' temporally correlated schemes rests on the untested assertion that symbol energy correlations dominate all other temporal correlations (phase, amplitude-phase, higher-order). The reported <5% NLI error is stated only for the energy-correlation case; without an ablation isolating non-energy contributions or a bound on their residual effect, the broader applicability claim is not supported by the validation.
  2. [Validation] Validation description: the experiments and simulations are described exclusively for energy-correlation scenarios. Because the model retains the Gaussian-noise approximation and omits other correlation terms by construction, the <5% error figure cannot be extrapolated to general temporally correlated modulation without additional evidence that the omitted terms remain negligible.
minor comments (1)
  1. [Abstract] The abstract states the premise that energy correlations 'are believed to exert the most significant influence' without citing prior literature or providing a short quantitative argument; adding a brief reference or order-of-magnitude estimate would strengthen the motivation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and the opportunity to clarify the scope and contributions of the MEGN model. We address the major comments point by point below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that MEGN provides a 'theoretical framework for analyzing and optimizing' temporally correlated schemes rests on the untested assertion that symbol energy correlations dominate all other temporal correlations (phase, amplitude-phase, higher-order). The reported <5% NLI error is stated only for the energy-correlation case; without an ablation isolating non-energy contributions or a bound on their residual effect, the broader applicability claim is not supported by the validation.

    Authors: We agree that the abstract phrasing could be read as implying broader applicability than intended. The manuscript explicitly notes that symbol energy correlations are 'believed to exert the most significant influence on NLI' and derives MEGN specifically as a memory extension of EGN to account for second-order energy correlations. The model omits other correlation types by construction. We will revise the abstract to state that the framework enables analysis and optimization for temporally correlated schemes via symbol energy correlations, while clarifying that non-energy terms are not addressed. revision: partial

  2. Referee: [Validation] Validation description: the experiments and simulations are described exclusively for energy-correlation scenarios. Because the model retains the Gaussian-noise approximation and omits other correlation terms by construction, the <5% error figure cannot be extrapolated to general temporally correlated modulation without additional evidence that the omitted terms remain negligible.

    Authors: The reported validation and <5% normalized average NLI power errors apply specifically to the energy-correlation scenarios examined in simulations and experiments, as these match the model's derivation. The Gaussian-noise approximation is retained and other correlation terms are omitted by design. We do not provide evidence or claim that omitted terms are negligible for arbitrary temporal correlations; the error metric is not extrapolated beyond the validated cases. The MEGN model supplies a concrete framework for the energy-correlation case, which is the primary focus given the stated belief in its dominance. revision: no

Circularity Check

0 steps flagged

MEGN derivation is a self-contained mathematical extension of EGN with independent validation

full rationale

The paper derives the MEGN model via explicit extension of the EGN model's moment-based NLI expressions to incorporate second-order symbol energy correlations across time, while retaining the Gaussian-noise approximation. This is presented as a direct mathematical construction from the EGN statistics plus correlation terms, validated separately by numerical simulations and experiments reporting normalized NLI errors below 5% over varied symbol rates and distances. No load-bearing step reduces by construction to a fitted parameter, self-citation chain, or input renaming; the central premise that energy correlations dominate is an explicit modeling choice whose scope is bounded by the reported validation rather than tautological.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The model rests on extending the Gaussian noise approximation with memory terms derived from higher-order moments; no free parameters are explicitly fitted in the abstract, but the dominance of energy correlations is asserted without independent proof.

axioms (2)
  • domain assumption Nonlinear interference power can be modeled via Gaussian statistics depending on second-, fourth-, and sixth-order moments of transmitted symbols.
    Core assumption carried over from the EGN model and extended to include correlations.
  • ad hoc to paper Symbol energy correlations are the dominant temporal correlation affecting NLI.
    Stated as believed to exert the most significant influence; used to justify focusing the memory extension on energy terms.
invented entities (1)
  • MEGN model no independent evidence
    purpose: Explicit accounting for symbol energy correlations in NLI estimation.
    New model introduced to handle non-i.i.d. transmissions.

pith-pipeline@v0.9.0 · 5542 in / 1318 out tokens · 39641 ms · 2026-05-10T15:06:41.061240+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

31 extracted references · 31 canonical work pages

  1. [1]

    Accounting for temporal energy correlations in the enhanced Gaussian noise model,

    K. Wu, G. Liga, M. Secondini, and A. Alvarado, “Accounting for temporal energy correlations in the enhanced Gaussian noise model,” inProc. Opt. Fiber Commun. Conf.San Francisco, CA, USA, Mar. 2025

    work page 2025

  2. [2]

    The GN-model of fiber non-linear propagation and its applications,

    P. Poggiolini, G. Bosco, A. Carena, V . Curri, Y . Jiang, and F. Forghieri, “The GN-model of fiber non-linear propagation and its applications,”J. Lightw. Technol., vol. 32, no. 4, pp. 694–721, Feb. 2014

    work page 2014

  3. [3]

    Properties of nonlinear noise in long, dispersion-uncompensated fiber links,

    R. Dar, M. Feder, A. Mecozzi, and M. Shtaif, “Properties of nonlinear noise in long, dispersion-uncompensated fiber links,”Opt. Express, vol. 21, no. 22, pp. 25 685–25 699, Nov. 2013

    work page 2013

  4. [4]

    EGN model of non-linear fiber propagation,

    A. Carena, G. Bosco, V . Curri, Y . Jiang, P. Poggiolini, and F. Forghieri, “EGN model of non-linear fiber propagation,”Opt. Express, vol. 22, no. 13, pp. 16 335–16 362, 2014

    work page 2014

  5. [5]

    On proba- bilistic shaping of quadrature amplitude modulation for the nonlinear fiber channel,

    T. Fehenberger, A. Alvarado, G. B ¨ocherer, and N. Hanik, “On proba- bilistic shaping of quadrature amplitude modulation for the nonlinear fiber channel,”J. Lightw. Technol., vol. 34, no. 21, pp. 5063–5073, Nov. 2016

    work page 2016

  6. [6]

    Kurtosis-limited sphere shaping for nonlinear interference noise reduction in optical channels,

    Y . C. G ¨ultekin, A. Alvarado, O. Vassilieva, I. Kim, P. Palacharla, C. Okonkwo, and F. M. Willems, “Kurtosis-limited sphere shaping for nonlinear interference noise reduction in optical channels,”J. Lightw. Technol., vol. 40, no. 1, pp. 101–112, Jan. 2021

    work page 2021

  7. [7]

    Extending fibre nonlinear interference power modelling to account for general dual- polarisation 4D modulation formats,

    G. Liga, A. Barreiro, H. Rabbani, and A. Alvarado, “Extending fibre nonlinear interference power modelling to account for general dual- polarisation 4D modulation formats,”Entropy, vol. 22, no. 11, p. 1324, Nov. 2020

    work page 2020

  8. [8]

    High- cardinality geometrical constellation shaping for the nonlinear fibre channel,

    E. Sillekens, G. Liga, D. Lavery, P. Bayvel, and R. I. Killey, “High- cardinality geometrical constellation shaping for the nonlinear fibre channel,”J. Lightw. Technol., vol. 40, no. 19, pp. 6374–6387, Aug. 2022

    work page 2022

  9. [9]

    On shaping gain in the nonlinear fiber-optic channel,

    R. Dar, M. Feder, A. Mecozzi, and M. Shtaif, “On shaping gain in the nonlinear fiber-optic channel,” inIEEE Int. Symp. Inform. Theory. Honolulu, HI, USA, June 2014, pp. 2794–2798

    work page 2014

  10. [10]

    Temporal probabilis- tic shaping for mitigation of nonlinearities in optical fiber systems,

    M. P. Yankov, K. J. Larsen, and S. Forchhammer, “Temporal probabilis- tic shaping for mitigation of nonlinearities in optical fiber systems,”J. Lightw. Technol., vol. 35, no. 10, pp. 1803–1810, May 2017

    work page 2017

  11. [11]

    Bandwidth efficient and rate-matched low-density parity-check coded modulation,

    G. B ¨ocherer, F. Steiner, and P. Schulte, “Bandwidth efficient and rate-matched low-density parity-check coded modulation,”IEEE Trans. Commun., vol. 63, no. 12, pp. 4651–4665, Dec. 2015

    work page 2015

  12. [12]

    Analysis of nonlinear fiber interactions for finite-length constant-composition sequences,

    T. Fehenberger, D. S. Millar, T. Koike-Akino, K. Kojima, K. Parsons, and H. Griesser, “Analysis of nonlinear fiber interactions for finite-length constant-composition sequences,”J. Lightw. Technol., vol. 38, no. 2, pp. 457–465, Jan. 2020

    work page 2020

  13. [13]

    Temporal energy analysis of symbol sequences for fiber nonlinear interference modelling via energy dispersion index,

    K. Wu, G. Liga, A. Sheikh, F. M. J. Willems, and A. Alvarado, “Temporal energy analysis of symbol sequences for fiber nonlinear interference modelling via energy dispersion index,”J. Lightw. Technol., vol. 39, no. 18, pp. 5766–5782, Sep. 2021

    work page 2021

  14. [14]

    A sequence selection bound for the capacity of the nonlinear fiber channel,

    S. Civelli, E. Forestieri, A. Lotsmanov, D. Razdoburdin, and M. Sec- ondini, “A sequence selection bound for the capacity of the nonlinear fiber channel,” inProc. Eur. Conf. Opt. Commun.Bordeaux, France, Sep. 2021

    work page 2021

  15. [15]

    List-encoding CCDM: A nonlinearity-tolerant shaper aided by energy dispersion index,

    K. Wu, G. Liga, A. Sheikh, Y . C. G ¨ultekin, F. M. Willems, and A. Alvarado, “List-encoding CCDM: A nonlinearity-tolerant shaper aided by energy dispersion index,”J. Lightw. Technol., vol. 40, no. 4, pp. 1064–1071, Feb. 2022

    work page 2022

  16. [16]

    Sequence-selection-based constellation shaping for nonlinear channels,

    S. Civelli, E. Forestieri, and M. Secondini, “Sequence-selection-based constellation shaping for nonlinear channels,”J. Lightw. Technol., vol. 42, no. 3, pp. 1031–1043, 2024

    work page 2024

  17. [17]

    Nonlinearity-tolerant temporal sign shaping for 64-QAM optical fiber transmissions,

    X. Han, K. Wu, F. Søren, and A. Alvarado, “Nonlinearity-tolerant temporal sign shaping for 64-QAM optical fiber transmissions,” inProc. Opt. Fiber Commun. Conf.Los Angeles, CA, USA, Mar. 2026

    work page 2026

  18. [18]

    Capacity of a nonlinear optical channel with finite memory,

    E. Agrell, A. Alvarado, G. Durisi, and M. Karlsson, “Capacity of a nonlinear optical channel with finite memory,”J. Lightw. Technol., vol. 32, no. 16, pp. 2862–2876, Aug. 2014

    work page 2014

  19. [19]

    Shaping lightwaves in time and frequency for optical fiber communication,

    J. Cho, X. Chen, G. Raybon, D. Che, E. Burrows, S. Olsson, and R. Tkach, “Shaping lightwaves in time and frequency for optical fiber communication,”Nature Commun., vol. 13, no. 1, p. 785, Feb. 2022

    work page 2022

  20. [20]

    On the nonlinear shaping gain with probabilistic shaping and carrier phase recovery,

    S. Civelli, E. Parente, E. Forestieri, and M. Secondini, “On the nonlinear shaping gain with probabilistic shaping and carrier phase recovery,”J. Lightw. Technol., vol. 41, no. 10, pp. 3046–3056, May 2023

    work page 2023

  21. [21]

    Predicting nonlinear interference for short-blocklength 4D probabilistic shaping,

    J. Deng, B. Chen, Z. Liang, Y . Lei, and G. Liga, “Predicting nonlinear interference for short-blocklength 4D probabilistic shaping,” inProc. Opt. Fiber Commun. Conf.San Francisco, CA, USA, Mar. 2025

    work page 2025

  22. [22]

    Probabilistic amplitude shaping and nonlinearity tolerance: Analysis and sequence selection method,

    M. T. Askari, L. Lampe, and J. Mitra, “Probabilistic amplitude shaping and nonlinearity tolerance: Analysis and sequence selection method,”J. Lightw. Technol., vol. 41, no. 17, pp. 5503–5517, Sep. 2023

    work page 2023

  23. [23]

    Constant composition distribution match- ing,

    P. Schulte and G. B ¨ocherer, “Constant composition distribution match- ing,”IEEE Trans. Inf. Theory, vol. 62, no. 1, pp. 430–434, Jan. 2016

    work page 2016

  24. [24]

    Huffman-coded sphere shaping for extended-reach single-span links,

    P. Skvortcov, I. Phillips, W. Forysiak, T. Koike-Akino, K. Kojima, K. Parsons, and D. S. Millar, “Huffman-coded sphere shaping for extended-reach single-span links,”IEEE J. Sel. Top. Quantum Electron., vol. 27, no. 3, pp. 1–15, May-June. 2021

    work page 2021

  25. [25]

    A Detailed Analytical Derivation of the GN Model of Non-Linear Interference in Coherent Optical Transmission Systems,

    P. Poggiolini, G. Bosco, A. Carena, V . Curri, Y . Jiang, and F. Forghieri, “A Detailed Analytical Derivation of the GN Model of Non-Linear Interference in Coherent Optical Transmission Systems,” arXiv, no. 1209.0394, pp. 1–24, 2012. [Online]. Available: http: //arxiv.org/abs/1209.0394

    work page arXiv 2012

  26. [26]

    On the accuracy of the GN-model an on analytical correction terms to improve it,

    A. Carena, G. Bosco, V . Curri, Y . Jiang, P. Poggiolini, and F. Forghieri, “On the accuracy of the GN-model an on analytical correction terms to improve it,”arXiv, 2014

    work page 2014

  27. [27]

    Exponentially- weighted energy dispersion index for the nonlinear interference anal- ysis of finite-blocklength shaping,

    K. Wu, G. Liga, Y . C. G ¨ultekin, and A. Alvarado, “Exponentially- weighted energy dispersion index for the nonlinear interference anal- ysis of finite-blocklength shaping,” inProc. Eur. Conf. Opt. Commun. Bordeaux, France, Sep. 2021

    work page 2021

  28. [28]

    Fast, flexible algorithm for calculating photon correlations,

    T. A. Laurence, S. Fore, and T. Huser, “Fast, flexible algorithm for calculating photon correlations,”Optics letters, vol. 31, no. 6, pp. 829– 831, 2006

    work page 2006

  29. [29]

    Binning method for artifact- free time-tag based correlation function calculations,

    O. Urquidi, J. Brazard, and T. B. Adachi, “Binning method for artifact- free time-tag based correlation function calculations,”Optics Letters, vol. 49, no. 16, pp. 4569–4572, 2024

    work page 2024

  30. [30]

    G. E. Box, G. M. Jenkins, G. C. Reinsel, and G. M. Ljung,Time series analysis: forecasting and control. John Wiley & Sons, 2015

    work page 2015

  31. [31]

    Mukherjee, I

    B. Mukherjee, I. Tomkos, M. Tornatore, P. Winzer, and Y . Zhao,Springer handbook of optical networks. Springer, 2015

    work page 2015