pith. sign in

arxiv: 2603.08411 · v2 · submitted 2026-03-09 · 💻 cs.IT · math.IT

Optical Communications with Relative Intensity Noise: Channel Modeling and Information Rates

Felipe Villenas , Yunus Can G\"ultekin , Alex Alvarado This is my paper

Pith reviewed 2026-05-15 13:48 UTC · model grok-4.3

classification 💻 cs.IT math.IT
keywords relative intensity noiseintensity modulation direct detectionchannel modelinggeneralized mutual informationmismatched decodingsignal-dependent noiseoptical communicationsachievable rates
0
0 comments X

The pith

In optical intensity-modulation links impaired by laser relative intensity noise, a memoryless mismatched decoder causes achievable rates to saturate as constellation size grows.

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

The paper starts from a continuous-time waveform model of an IM/DD optical channel and derives a discrete-time equivalent in which RIN produces signal-dependent noise whose variance depends polynomially on the current symbol and carries memory from prior symbols. Achievable rates are then computed via the generalized mutual information under a practical memoryless decoding metric. Numerical evaluation shows that GMI stops increasing once constellations become dense, because individual symbols contribute to the rate in a nonsymmetric and nonvanishing way. A reader would care because the result indicates that simply raising modulation order brings no further rate benefit once the receiver ignores channel memory, a common engineering choice.

Core claim

The derived discrete-time channel model exhibits signal-dependent noise with memory whose conditional variance is a polynomial function of the transmitted symbol. When a memoryless mismatched decoder is used, the generalized mutual information saturates with growing constellation cardinality; the saturation arises from the nonsymmetric, nonvanishing contributions that different symbols make to the GMI.

What carries the argument

Discrete-time channel model whose noise variance is a polynomial function of the current symbol and carries memory from previous symbols, induced by relative intensity noise.

Load-bearing premise

The continuous-time waveform model together with the chosen sampling and filtering steps fully captures all relevant RIN dynamics, so that the resulting discrete-time polynomial variance expression remains accurate.

What would settle it

A physical or high-fidelity simulation experiment in which constellation size is increased while keeping the same memoryless decoder and measuring whether GMI indeed plateaus rather than continuing to rise.

Figures

Figures reproduced from arXiv: 2603.08411 by Alex Alvarado, Felipe Villenas, Yunus Can G\"ultekin.

Figure 1
Figure 1. Figure 1: Equivalent high-speed IM-DD system with laser RIN and external modulation. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Histogram of the samples Yk. The contributions of the thermal noise Qk and signal-dependent noise Zk are also plotted independently. of the transmitted symbol Ak = an: the larger the value of an, the larger the variance. Motivated by the results in Example 1, we are now in￾terested in the conditional first and second moments of the noise samples Zk, conditioned on a given transmitted symbol Ak = an. Using … view at source ↗
Figure 3
Figure 3. Figure 3: Conditional variance E [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: GMI vs. OMA for the system in Fig. 1 using the parameters in Table I. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: GMI at OMA = 25 dBm vs. M. The GMI converges as M → ∞ (dashed lines). The maximum value is obtained for M = 16. (OMA = 25 dBm). This figure shows that the saturated GMI has a maximum at M = 16, and confirms that using values of M larger than M = 8 provides little or no gain in terms of GMI. As mentioned above, we believe that using a mismatched metric with memory would remove this saturation. To conclude, … view at source ↗
read the original abstract

We consider optical communications with intensity modulation and direct detection affected by laser relative intensity noise (RIN). Starting from a continuous-time waveform model, we derive an equivalent discrete-time channel model. As a result of RIN, the resulting channel model exhibits signal-dependent noise with memory. Unlike the commonly-assumed model in the literature, the conditional variance of this noise term has a polynomial dependence on the symbol of interest. Finally, we study achievable information rates for this channel under practically-relevant system parameters. We take a mismatched decoding approach and compute the generalized mutual information (GMI) using a memoryless decoding metric. Our numerical results show that when the memory in the channel is ignored by the receiver, GMI saturates as the constellation size increases, and thus, dense constellations do not offer gains. We also show that this saturation results from nonsymmetric nonvanishing contributions of the symbols to the GMI.

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 / 3 minor

Summary. The paper starts from a continuous-time waveform model of an intensity-modulation direct-detection optical channel impaired by laser relative intensity noise (RIN). It derives an equivalent discrete-time channel whose noise is signal-dependent, possesses memory, and has conditional variance that is a polynomial function of the transmitted symbol. Achievable rates are evaluated via generalized mutual information (GMI) under a mismatched memoryless decoding metric. Numerical results for practical parameters show that GMI saturates with increasing constellation cardinality; the authors attribute the saturation to nonsymmetric, nonvanishing per-symbol contributions to the GMI.

Significance. If the continuous-to-discrete derivation holds, the work supplies a more accurate RIN channel model than the memoryless or linear-variance approximations common in the literature. The saturation result supplies a concrete, falsifiable prediction for system designers: beyond a modest cardinality, denser constellations yield diminishing returns when the receiver treats the channel as memoryless. The explicit polynomial-variance structure and the GMI analysis constitute reusable modeling tools for RIN-limited links.

major comments (2)
  1. [§3] §3 (discrete-time model derivation): the claim that the conditional variance is exactly polynomial in the symbol rests on the specific sampling instant and front-end filter response; the manuscript should state the filter impulse response and sampling phase explicitly so that the polynomial coefficients can be reproduced from the continuous-time parameters.
  2. [§4] §4 (GMI evaluation): the saturation result is obtained under a memoryless metric; the paper should report the GMI gap to a metric that retains the derived memory (e.g., via a trellis or factor-graph decoder) for at least one constellation size, to confirm that the observed saturation is not an artifact of the memoryless restriction.
minor comments (3)
  1. [Table I] Table I: the RIN strength parameter is listed without units; add the physical unit (e.g., 1/Hz) for clarity.
  2. [Figure 3] Figure 3 caption: the legend entries for different constellation sizes are difficult to distinguish in grayscale; use distinct line styles or markers.
  3. [Eq. (12)] Eq. (12): the definition of the mismatched metric should explicitly indicate that it ignores the memory term derived in Eq. (9).

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading, positive assessment of the work's significance, and constructive suggestions. We address each major comment below and will revise the manuscript to improve clarity and reproducibility.

read point-by-point responses

  1. Referee: [§3] §3 (discrete-time model derivation): the claim that the conditional variance is exactly polynomial in the symbol rests on the specific sampling instant and front-end filter response; the manuscript should state the filter impulse response and sampling phase explicitly so that the polynomial coefficients can be reproduced from the continuous-time parameters.

    Authors: We agree that these details are necessary for full reproducibility. In the revised manuscript we will explicitly state that the front-end filter is a rectangular integrate-and-dump filter with impulse response equal to the symbol duration and that sampling occurs at the center of each symbol interval. With these choices the conditional variance reduces exactly to the reported polynomial in the transmitted symbol; the coefficients can then be obtained directly from the continuous-time RIN parameters via the derivation in §3. revision: yes

  2. Referee: [§4] §4 (GMI evaluation): the saturation result is obtained under a memoryless metric; the paper should report the GMI gap to a metric that retains the derived memory (e.g., via a trellis or factor-graph decoder) for at least one constellation size, to confirm that the observed saturation is not an artifact of the memoryless restriction.

    Authors: The paper deliberately evaluates GMI under the memoryless mismatched metric because this corresponds to the low-complexity decoding used in practical high-speed IM/DD systems. The observed saturation is a direct consequence of this mismatch: the memory in the RIN-induced noise produces nonsymmetric, nonvanishing per-symbol contributions to the GMI, which we quantify explicitly. We will add a clarifying paragraph in §4 emphasizing that the saturation is an artifact of the memoryless restriction and is therefore expected; a quantitative gap to a memory-retaining decoder (e.g., BCJR) is left for future work, as implementing such a decoder lies outside the scope of the present contribution focused on practical decoding. revision: partial

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper starts from a standard continuous-time optical waveform model for RIN-affected intensity modulation and derives the discrete-time channel with signal-dependent polynomial-variance noise via sampling and filtering. The GMI saturation result under memoryless mismatched decoding is obtained by direct numerical evaluation of the resulting model; no step reduces by construction to a fitted parameter, self-definition, or load-bearing self-citation. The derivation remains self-contained against external benchmarks and does not invoke uniqueness theorems or ansatzes from prior author work.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claims rest on a standard continuous-time optical waveform model and the validity of the discretization step; no new entities are postulated and free parameters are limited to system-level RIN strength.

free parameters (1)
  • RIN strength parameter
    The relative intensity noise level is treated as an input parameter that scales the polynomial variance term in the derived model.
axioms (1)
  • domain assumption Optical intensity-modulated signals can be represented by a continuous-time waveform whose power fluctuations are dominated by laser RIN.
    Invoked at the start of the derivation to obtain the discrete-time channel statistics.

pith-pipeline@v0.9.0 · 5455 in / 1281 out tokens · 60694 ms · 2026-05-15T13:48:45.213440+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

18 extracted references · 18 canonical work pages

  1. [1]

    Beyond 1 Tb/s intra-data center interconnect technology: IM-DD or coherent?

    X. Zhou, R. Urata, and H. Liu, “Beyond 1 Tb/s intra-data center interconnect technology: IM-DD or coherent?”J. Lightw. Technol., vol. 38, no. 2, pp. 475–484, Jan. 2020

    work page 2020

  2. [2]

    Modulation format and digital signal processing for IM-DD optics at post-200G era,

    D. Che and X. Chen, “Modulation format and digital signal processing for IM-DD optics at post-200G era,”J. Lightw. Technol., vol. 42, no. 2, pp. 588–605, Jan. 2024

    work page 2024

  3. [3]

    Experimental demonstration of advanced modulation formats for data center networks on 200 Gb/s lane rate IMDD links,

    J. Wei, T. Rahman, S. Calabr `o, N. Stojanovic, L. Zhang, C. Xie, Z. Ye, and M. Kuschnerov, “Experimental demonstration of advanced modulation formats for data center networks on 200 Gb/s lane rate IMDD links,”Optics Express, vol. 28, no. 23, pp. 35 240–35 250, Nov. 2020

    work page 2020

  4. [4]

    Baseline proposals for 200G/L PMD specifications for single wavelength 500 m and 2 km standards,

    B. Welch, J. Ingham, E. Bernier, and P. Dawe, “Baseline proposals for 200G/L PMD specifications for single wavelength 500 m and 2 km standards,”IEEE P802.3dj Ethernet Task Force, Feb. 2023

    work page 2023

  5. [5]

    G. P. Agrawal and N. K. Dutta,Semiconductor lasers. Springer Science & Business Media, 2013

    work page 2013

  6. [6]

    Keysight Technologies,Digital Communication Analyzer (DCA), Mea- sure Relative Intensity Noise (RIN), July 2014

    work page 2014

  7. [7]

    4-PAM for high-speed short-range optical communications,

    K. Szczerba, P. Westbergh, J. Karout, J. S. Gustavsson, ˚A. Haglund, M. Karlsson, P. A. Andrekson, E. Agrell, and A. Larsson, “4-PAM for high-speed short-range optical communications,”J. Opt. Commun. Netw., vol. 4, no. 11, pp. 885–894, Oct. 2012

    work page 2012

  8. [8]

    An analyt- ical model for performance estimation in modern high-capacity IMDD systems,

    G. Rizzelli, P. Torres-Ferrera, F. Forghieri, and R. Gaudino, “An analyt- ical model for performance estimation in modern high-capacity IMDD systems,”J. Lightw. Technol., vol. 42, no. 5, pp. 1443–1452, Mar. 2023

    work page 2023

  9. [9]

    Reach extension of a 53.12 Gbps PAM-4 IM/DD signals with optimized extinction ratio in amplified multi-span O-Band transmission,

    R. Maneekut, D. J. Elson, Y . Wakayama, N. Yoshikane, and P. Kaew- plung, “Reach extension of a 53.12 Gbps PAM-4 IM/DD signals with optimized extinction ratio in amplified multi-span O-Band transmission,” IEEE Access, vol. 12, pp. 34 656–34 667, Mar. 2024

    work page 2024

  10. [10]

    Hui,Introduction to Fiber-Optic Communications

    R. Hui,Introduction to Fiber-Optic Communications. Academic Press, 2019

    work page 2019

  11. [11]

    On the achievable information rates of finite state ISI channels,

    H. D. Pfister, J. B. Soriaga, and P. H. Siegel, “On the achievable information rates of finite state ISI channels,” inProc. IEEE Global Commun. Conf. (GLOBECOM), San Antonio, TX, USA, Aug. 2001

    work page 2001

  12. [12]

    Information rates and error exponents of compound channels with application to antipodal signaling in a fading environment,

    G. Kaplan and S. Shamai (Shitz), “Information rates and error exponents of compound channels with application to antipodal signaling in a fading environment,”A ¨EU, vol. 47, no. 4, pp. 228–239, 1993

    work page 1993

  13. [13]

    On information rates for mismatched decoders,

    N. Merhav, G. Kaplan, A. Lapidoth, and S. Shamai (Shitz), “On information rates for mismatched decoders,”IEEE Trans. on Inf. Theory, vol. 40, no. 6, pp. 1953–1967, Nov. 1994

    work page 1953

  14. [14]

    Bit-interleaved coded modulation revisited: A mismatched decoding perspective,

    A. Martinez, Guill ´en i F `abregas, G. Caire, and F. M. J. Willems, “Bit-interleaved coded modulation revisited: A mismatched decoding perspective,”IEEE Trans. Inf. Theory, vol. 55, no. 6, pp. 2756–2765, June 2009

    work page 2009

  15. [15]

    Szczecinski and A

    L. Szczecinski and A. Alvarado,Bit-Interleaved Coded Modulation: Fundamentals, Analysis, and Design. Chichester, UK: John Wiley & Sons, 2015

    work page 2015

  16. [16]

    Achievable information rates for fiber optics: Applications and computations,

    A. Alvarado, T. Fehenberger, B. Chen, and F. M. J. Willems, “Achievable information rates for fiber optics: Applications and computations,”J. Lightw. Technol., vol. 36, no. 2, pp. 424–439, Jan. 2018

    work page 2018

  17. [17]

    200 Gbps/lane IM/DD technolo- gies for short reach optical interconnects,

    X. Pang, O. Ozolins, R. Lin, L. Zhang, A. Udalcovs, L. Xue, R. Schatz, U. Westergren, S. Xiao, W. Huet al., “200 Gbps/lane IM/DD technolo- gies for short reach optical interconnects,”J. Lightw. Technol., vol. 38, no. 2, pp. 492–503, Jan. 2020

    work page 2020

  18. [18]

    High-SNR asymp- totics of mutual information for discrete constellations with applications to BICM,

    A. Alvarado, F. Br ¨annstr¨om, E. Agrell, and T. Koch, “High-SNR asymp- totics of mutual information for discrete constellations with applications to BICM,”IEEE Trans. Inf. Theory, vol. 60, no. 2, pp. 1061–1076, Feb. 2014

    work page 2014