Pith. sign in

REVIEW 2 major objections 5 minor 39 references

Transition-Aware Routing in Hybrid Hollow-Core/Single-Mode Fiber Networks: A Cost--Throughput Investigation

T0 review · 2 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

Pith's one-line read In hybrid hollow-core/single-mode fiber networks, the best protected routing scheme is determined by a single operator parameter: the exchange rate κ between transitions avoided and throughput lost.

desk verdict A carefully run simulation study whose two new middle-ground schemes and exchange-rate framing are useful, but the decision-rule breakpoints rest on an unvalidated beta=0.60 in the dominant blocking constraint and need a sensitivity sweep before they are quoted. read the letter →

arxiv 2607.14324 v1 pith:5INZMGVV submitted 2026-07-15 cs.NI physics.optics

classification cs.NIphysics.optics
keywords hollow-corefibersingle-moderoutingandspectrumassignmentGSNRtransitionpenaltydynamicprovisioningprotectedlightpathsdecisionrule
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tackles a routing problem that appears only when a network mixes hollow-core fiber (HCF) with standard single-mode fiber (SMF): every HCF–SMF handover costs physical-layer margin, but avoiding handovers forces detours that waste spectrum. Using an event-driven simulator with a per-transition GSNR (generalized signal-to-noise ratio) penalty, six protected routing schemes are compared across six topologies and five HCF deployment fractions. Two newly introduced middle-ground schemes—GMR-T, which reranks GSNR-optimal candidates by transition count, and BD-TPAR, which bounds detours to 20%—cut transitions by roughly 22% and 11% at only 3% and 1% throughput cost, while the aggressive minimizers TPAR (transition-penalty-aware routing) and GFJ (joint GSNR-fiber-transition scheme) halve transitions at a 20–25% cost. The paper's key practical output is a piecewise decision rule mapping an operator's transition-to-throughput exchange rate κ to one of five schemes, and the finding that contiguous HCF deployment reduces transitions by ~40% on average while improving carried traffic.

What carries the argument

The state-augmented graph used by TPAR and GFJ: each physical node is duplicated as (node, incoming-fiber-type), so an edge cost becomes link length (or inverse GSNR) plus a transition penalty matrix P[HCF→SMF, SMF→HCF] that charges every fiber-type handover. Around this, the decision rule is the upper envelope of six utility lines Us = ws + κ(n̄DA-RSA − n̄s), where ws is carried-traffic lift over DA-RSA and n̄s is mean transitions per accepted demand; the crossing points of those lines define the five κ regions.

What would settle it

Measure the differential-delay tolerance of a production 1+1 coherent transponder on a hybrid HCF/SMF link: if the receiver can realign working and protection copies with delay asymmetry larger than the paper's Δτmax (β=0.60) without bit errors, then the HC1 blocking fraction in Fig. 7 should fall and the carried-traffic penalty of TPAR/GFJ should drop below 20–25%, invalidating the κ thresholds that depend on it.

Watch

Extended reading notes

Core claim

The central claim is that transition-aware routing in hybrid HCF/SMF networks is not a binary choice — the right scheme depends on how much a transition costs the operator relative to a unit of carried traffic. Concretely, the paper models each HCF↔SMF crossing with a calibrated 0.4 dB GSNR penalty and a hard latency-asymmetry constraint, then shows that pure transition minimizers (TPAR, GFJ) halve the mean transition count but sacrifice 20–25% of carried traffic, mostly because they trip the latency-asymmetry gate (HC1). The introduced middle-ground schemes avoid most of that penalty: BD-TPAR (−11% transitions, −1.2% throughput) and GMR-T (−22%, −3%). These numbers feed a linear utility mod

Load-bearing premise

The central trade-off's magnitude rests on the hard latency-asymmetry limit Δτmax = 0.60 × mean shortest-path length × (1/vg,SMF − 1/vg,HCF), which is assumed to model receiver buffer capability and drives 69–75% of blocking at 300 Erlang; if real transponders allow more than 60% of that worst-case delay gap, the throughput penalty of transition-minimizing schemes shrinks and the decision-rule thresholds change.

Editorial extensions

If this is right

  • An operator needs to estimate only one number, κ, to choose among DA-RSA, BD-TPAR, GMR-T, TPAR, and GFJ; no per-topology recalibration is required for the choice.
  • Under fragmented (random) HCF rollout, BD-TPAR is the practical default at ~11% fewer transitions for ~1% throughput cost, with GMR-T as a lower-complexity fallback offering ~22% for ~3%.
  • Contiguous HCF deployment is a stronger lever than any routing scheme, cutting transitions by 23–54% (mean ~42%) while raising carried traffic, so rollout planning should prioritize geographic clustering before tuning routing.
  • The scheme ordering and κ-based rule persist in the L-band under a band-averaged CO2 absorption penalty, so the decision applies across both C- and L-band transmission.
  • At heavy load, latency-asymmetry rejection (HC1), not spectrum scarcity, is the dominant blocking mechanism, implying that hardware improvements in receiver differential-delay buffering would reshape the throughput trade-off landscape.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The upper-envelope decision method is not specific to fiber transitions: the same utility-line construction could price other additive per-route costs (security checkpoints, regulatory boundaries, maintenance-heavy segments) in any network.
  • Because the 20–25% penalty of TPAR/GFJ is driven by the assumed receiver buffer budget β=0.60 in Eq. (3), a transponder with a larger differential-delay tolerance would likely rescue much of that throughput loss and shift the κ thresholds downward; this is testable against public DSP buffer specifications.
  • The small (~4%) transition reduction of pure GMR suggests that GSNR awareness alone does not avoid fiber crossings, so an operator running GSNR-only routing in a hybrid plant is systematically leaving transition savings unclaimed — explicit transition terms are needed.
  • Contiguous deployment achieves transition reductions comparable to the most aggressive routing schemes without their throughput penalty, implying deployment order and routing weight could be co-optimized to reach the same gain at lower routing complexity.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper studies routing in hybrid hollow-core/single-mode fiber networks, where each HCF-SMF transition carries a physical-layer penalty. Using a common event-driven simulator with paired seeds and 95% confidence intervals, it compares six protected routing schemes over six topologies, five HCF deployment fractions, and dynamic loads. The central findings are that aggressive transition-minimizing schemes (TPAR, GFJ) roughly halve per-demand transitions at a 20-25% carried-traffic penalty; intermediate schemes (GMR-T, BD-TPAR) offer smaller transition reductions at much lower cost; and a decision rule based on an operator-supplied transition-to-throughput exchange rate κ selects among schemes. The paper also presents L-band and availability cross-checks and a sensitivity sweep of the per-transition GSNR penalty ηtrans.

Significance. If the results hold, the paper provides a useful, directly actionable rule for operators choosing among transition-aware routing schemes in hybrid fiber networks. The strengths of the work are its controlled experimental design: all schemes share the same seeds, topologies, and demand realizations; the block-reason decomposition identifies the dominant blocking mechanism; the ηtrans sweep and the L-band/availability cross-checks test the sensitivity of the main conclusions; and the decision rule is a clear, falsifiable output. The paper is honest about the exploratory nature of several modeling choices, and it explicitly distinguishes measured parameters from conservative estimates.

major comments (2)
  1. [Section 2.D, Eq. (3); Section 5.F, Fig. 7; Section 6.B] The decision rule of Eq. (9) and the stated κ breakpoints depend on the carried-traffic penalties of TPAR/GFJ, which the paper attributes primarily to the latency-asymmetry hard constraint HC1. Eq. (3) fixes Δτmax = 0.60 × mean shortest-path length × group-delay gap, and the constant β=0.60 is introduced without derivation, citation, or a companion sweep. Fig. 7 shows that HC1 accounts for 69-75% of rejections on CORONET, so the magnitude of the TPAR/GFJ penalty is directly set by an unvalidated constant. A larger differential-delay buffer would relax HC1 and likely reduce the TPAR/GFJ penalty, shifting the κ≥40 and κ≥77 thresholds downward; a smaller buffer would do the opposite. The ηtrans sweep does not cover this concern because ηtrans affects feasibility uniformly while β changes the dominant blocking mechanism that separates the aggressive minimizers from the baselines. Please just
  2. [Section 6.B; Fig. 3(c)] The breakpoints κ=9, 16, 40, and 77 are computed from the point estimates ws and n̄s in Fig. 3, which are six-topology means at a single operating point (50% HCF, 300 Erlang). The carried-traffic differences among DA-RSA, BD-TPAR, and GMR-T are small (about -1.2% to -3.5%), and the paper itself states that per-topology differences among the four leading schemes are modest at 95% confidence. The breakpoints that separate DA-RSA/BD-TPAR/GMR-T are therefore likely to be sensitive to seed noise and to the choice of averaging, yet no uncertainty propagation or sensitivity analysis of the breakpoints is reported. At minimum, report confidence intervals for the crossing points or show that the region boundaries are robust across topologies rather than relying on pooled means.
minor comments (5)
  1. [Throughout] Several internal cross-references appear to point to non-existent or ambiguous sections: e.g., Section 2.C refers to the ηtrans sweep as 'Section E' when the sweep appears in Section 5.E; Section 5.B refers to 'Section D' for contiguous rollout; Section 5.F is sometimes called 'Section F' in the text. The manuscript would benefit from consistent, resolved section numbering.
  2. [Section 2.C] There is an unresolved placeholder reference in the text: 'gain-control loops must suppress [?]'. If this is meant to be a citation, it should be filled in before publication.
  3. [Fig. 7] The y-axis label 'Block-reason rate' is ambiguous: the four stacked components appear to sum to the total blocking probability, but the reader has to infer this. Please clarify whether these are rates as a fraction of all demands or as a fraction of blocked demands, and add the total blocking probability to the figure or caption.
  4. [Section 6.B] The light-load claim that middle-ground schemes edge ahead of DA-RSA by +2 to +3% at 50-100 Erlang is not shown in a figure. A corresponding plot or at least a table would make the claim verifiable.
  5. [Data availability] The data availability statement indicates that data are not publicly available. Given the paper's contribution is a decision rule derived from simulation, releasing the simulator or the per-seed results would substantially strengthen reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: predictions are simulator outputs and the decision rule is an explicit upper envelope, not a fitted reconstruction.

full rationale

The paper's derivation chain is simulation-based, and the central predictions (transition reductions and carried-traffic penalties) are direct outputs of a common event-driven simulator, not algebraic consequences of the model equations. The per-transition penalty ηtrans is an explicitly calibrated input, but it is applied symmetrically to all schemes and swept in Section E; the ranking is reported insensitive to it, so it is not a fitted parameter that reconstructs the result. HC1's β=0.60 is an unvalidated modeling constant, but that is a parameter-uncertainty/correctness concern, not circularity: it is an input assumption, and the paper does not claim to derive it from the target results. The decision rule Eq. (9) is by construction the upper envelope of the six measured (ws, nbar) points; the paper candidly states that it reads directly off Fig. 3, so no prediction is disguised as a fit. Self-citations [1], [18], [26] provide context or supporting capacity maps; the load-bearing group-delay gap is computed from Table 1's ng values (1.468 vs 1.0003), and no uniqueness theorem or ansatz is imported from the authors' prior work. The availability cross-check explicitly shows that its exploratory splice term drives no conclusion. No step reduces, by definition or by self-citation, to its own inputs.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central claims rest on a simulation model with several hand-set parameters and standard domain assumptions. The qualitative ranking of the six schemes is reported as insensitive to the per-transition penalty ηtrans, but the quantitative magnitudes (1%, 3%, 20-25%) are specific to the listed operating points. The availability model is explicitly exploratory.

free parameters (7)
  • ηtrans (per-transition GSNR penalty) = 0.4 dB
    Order-of-magnitude estimate from 0.3 dB splice/coupler loss and 0.1 dB EDFA transient; the paper calls it a calibrated sensitivity parameter, not a measured constant; applied in Eq. (2) to all schemes; headline numbers are at 0.4 dB.
  • Transition-penalty matrix P off-diagonal (H→S, S→H) = 1.00, 0.89
    Hand-set direction asymmetry in Eq. (7); not swept; affects TPAR/GFJ path selection.
  • GFJ transition weight λGFJ = 0.3
    Tunable scalar in Eq. (8); placed in the saturated regime; not swept; the abstract's GFJ results are at this value.
  • BD-TPAR detour cap δ = 1.2
    Detour budget in Eq. (6); the '11% cut at 1% cost' result for BD-TPAR depends on this; only endpoints δ=1 and δ=∞ are discussed, no sweep.
  • Latency-asymmetry factor β = 0.60
    Eq. (3); scales the delay-asymmetry buffer; HC1 dominates blocking (69-75%), so this constant directly shapes the TPAR/GFJ throughput penalty; not swept or supported by buffer specs.
  • Occupancy-state breakpoints = one third and two thirds of 80 channels (with hysteresis)
    Hand-chosen thresholds for LOW/MED/HIGH states in Section 2.B; not reported exactly, not swept.
  • Splice failure rate λspl = 1×10⁻⁵ h⁻¹
    Exploratory availability parameter; the authors state it drives no conclusion since crossover requires ~85× this rate.
assumptions (6)
  • domain assumption Extended GN model with incoherent accumulation (Eq. 1) is an adequate physical-layer model.
    Used for GSNR estimation; standard in the field (Poggiolini, Filer), extended with an IMI term; the paper does not validate against experimental transmissions.
  • domain assumption Per-span GSNR invariance under locally-optimized launch power lets the simulator scale one reference span by link length.
    Section 2.A states 'Since LOGO makes the per-span GSNR invariant along a uniform-fiber link, we accumulate inverse-GSNR by scaling one reference span's value by the link length rather than summing spans individually.' This is a modeling simplification.
  • domain assumption Occupancy-state discretization (LOW/MED/HIGH) with precomputed GSNR approximates per-arrival interference.
    Section 2.B: breakpoints near one third and two thirds of 80 channels, with hysteresis; no validation against continuous recomputation.
  • domain assumption Random uniform and contiguous BFS deployments represent the two relevant rollout patterns.
    Section 2.A; real-world deployments may differ; contiguous pattern is described as likely closer to real rollouts.
  • domain assumption 1+1 protection with hot standby and pre-converged DSP is a valid operational model.
    Section 2.C excludes DSP reconvergence citing [33].
  • domain assumption Poisson arrivals, exponential holding times, and uniform source-destination pairs model dynamic traffic.
    Section 2.D; 3000 requests per trial, 10 seeds; results at 300 Erlang.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Transition-Aware Routing in Hybrid Hollow-Core/Single-Mode Fiber Networks: A Cost--Throughput Investigation." pith.science (2026). https://pith.science/paper/5INZMGVV

@misc{pith2026260714324,
  author       = {Pith},
  title        = {Pith review of: Transition-Aware Routing in Hybrid Hollow-Core/Single-Mode Fiber Networks: A Cost--Throughput Investigation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5INZMGVV}},
  note         = {Machine review of arXiv:2607.14324}
}
read the original abstract

Incremental deployment of hollow-core fiber (HCF) in single-mode-fiber (SMF) networks introduces a routing tradeoff: reducing HCF-SMF transitions can improve physical-layer feasibility, but overly transition-averse routing incurs harmful path detours. We study this tradeoff using a common event-driven simulator that compares six protected routing schemes spanning fiber-blind, generalized signal-to-noise ratio (GSNR)-aware, and explicitly transition-aware designs on hybrid HCF/SMF topologies. The model includes a per-transition GSNR penalty and an exploratory splice-failure availability term. Across six reference topologies, five HCF deployment fractions, and dynamic loads at 300 Erlang, the strongest transition minimizers, transition-penalty-aware routing (TPAR) and the GSNR/fiber-transition joint scheme (GFJ), halve the mean transition count at a 20-25% carried-traffic penalty. Among the intermediate designs, GSNR- maximal routing with transition-aware reranking (GMR-T) cuts transitions by approximately 22% relative to distance-adaptive routing and spectrum assignment (DA-RSA) at a 3% throughput cost, while bounded-detour TPAR (BD-TPAR) cuts transitions by approximately 11% at only a 1% cost. Deployment pattern also matters: contiguous HCF rollout lowers transitions by approximately 40% on average while improving carried traffic, reducing the benefit of aggressive transition-aware routing. These results support BD-TPAR as a practical default under fragmented deployment, GMR-T as a lower-complexity alternative, and TPAR or GFJ only where the external cost of transitions is high.

Figures

Figures reproduced from arXiv: 2607.14324 by the authors.

Figure 1
Figure 1. Overview. (a) A 1+1-protected demand traverses HCF and SMF segments; each HCF↔SMF transition (η) adds splice loss and an EDFA gain transient to the Extended-GSNR feasibility budget [Eq. (2)] and an exploratory splice-failure term to the avail￾ability model (Section E). (b) CORONET fiber-type map at 50% HCF under uniform-random deployment (one seed); the interleav￾ing of HCF (blue solid) and SMF (red dashed) drives t… view at source ↗
Figure 2
Figure 2. Blocking probability vs. Erlang load at 50% HCF; bands are 95% confidence intervals over ten seeds. NSFNET), because its shortest physical paths claim wavelengths most efficiently when spectrum is tight. The middle-ground group (GMR-T, BD-TPAR) tracks DA-RSA closely—their rerank￾ing and bounded detour keep path lengths near the shortest feasible option. The transition-minimizers (TPAR, GFJ) block substantially more … view at source ↗
Figure 3
Figure 3. The transition/throughput trade-off at 50% HCF, 300 Erlang. (a) Modulation mix per scheme (mean over six topologies). (b) Cross-fiber transitions per accepted demand (bars: six-topology mean; whiskers: spread across topologies), annotated with the reduction relative to DA-RSA. (c) Carried-traffic lift over DA-RSA per scheme and topology. 0.00 0.25 0.50 0.75 1.00 HCF fraction 10 1 10 0 2 × 10 1 3 × 10 1 4 × 10 1 6 × … view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Blocking probability, carried traffic, p99 latency asym￾metry, and cross-fiber transition count vs. HCF fraction (CORONET, 300 Erlang). penalty; the transition-minimizers’ advantage shrinks under contiguous rollout, so the decision rule applies most strongly on the fra…
Figure 5
Figure 5. Figure 5: Cross-fiber transitions per accepted demand (left) and carried traffic (right) under uniform-random (solid) vs. contiguous (hatched) HCF deployment (pHCF = 0.5, 300 Erlang). 0 1 trans [dB] 0.5 0.6 0.7 Blocking probability 0 1 trans [dB] 20 0 Lift vs DA-RSA [%] DA-RSA G…
Figure 6
Figure 6. Figure 6: Blocking probability (left) and carried-traffic lift over DA-RSA (right) vs. the per-transition penalty ηtrans (CORO￾NET, 50% HCF, 300 Erlang). DA-RSA GMR GMR-T BD-TPAR TPAR GFJ 0.0 0.2 0.4 0.6 Block-reason rate latency-asymmetry (HC1) no-working-path GSNR-infeasible (…
Figure 7
Figure 7. Figure 7: Block-reason rates per scheme (CORONET, 50% HCF, 300 Erlang). GBd signal bandwidth rather than sampled at the carrier. This is a first-order OSNR treatment: the EDFA compensates the band-averaged excess loss and consequently raises ASE. It is a routing-layer penalty ma…
Figure 8
Figure 8. Figure 8: L-band results with per-channel CO2 penalties. (a) Blocking per topology at 50% HCF and 300 Erlang; the scheme ordering matches the C-band. (b) Carried traffic versus HCF fraction on CORONET; traffic increases with HCF deployment. 0.00 0.25 0.50 0.75 1.00 HCF fraction …
Figure 9
Figure 9. Figure 9: Composite 1+1 availability on CORONET at 300 Er￾lang. (a) Mean composite availability versus HCF deployment fraction. (b) Mean composite availability versus the per-splice failure rate at 50% HCF, recomputed over the captured paths of accepted demands without rerouting…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 1 linked inside Pith

  1. [1]

    Hollow-core and standard single-mode fiber hybrid optical networks: A multi-topology investigation of protection switching,

    M. G. Saber and Z. Jiang, “Hollow-core and standard single-mode fiber hybrid optical networks: A multi-topology investigation of protection switching,” inThe 52nd European Conference on Optical Communica- tion (ECOC 2026),(Malaga, Spain, 2026)

  2. [2]

    Finding theK shortest loopless paths in a network,

    J. Y . Y en, “Finding theK shortest loopless paths in a network,” Manag. Sci.17, 712–716 (1971)

  3. [3]

    Disjoint paths in a network,

    J. W. Suurballe, “Disjoint paths in a network,” Networks4, 125–145 (1974)

  4. [4]

    Routing and spectrum allo- cation in elastic optical networks: A tutorial,

    B. C. Chatterjee, N. Sarma, and E. Oki, “Routing and spectrum allo- cation in elastic optical networks: A tutorial,” IEEE Commun. Surv. & Tutorials17, 1776–1800 (2015)

  5. [5]

    Survivable WDM mesh networks,

    S. Ramamurthy, L. Sahasrabuddhe, and B. Mukherjee, “Survivable WDM mesh networks,” J. Light. Technol.21, 870–883 (2003)

  6. [6]

    A survey on physical layer impairments aware routing and wavelength assignment algo- rithms in optical networks,

    S. Azodolmolky, M. Klinkowski, E. Marín,et al., “A survey on physical layer impairments aware routing and wavelength assignment algo- rithms in optical networks,” Comput. Networks53, 926–944 (2009)

  7. [7]

    The GN model of non-linear propagation in uncompen- sated coherent optical systems,

    P . Poggiolini, “The GN model of non-linear propagation in uncompen- sated coherent optical systems,” J. Light. Technol.30, 3857–3879 (2012)

  8. [8]

    Multi-vendor experimental validation of an open source QoT estimator for optical networks,

    M. Filer, M. Cantono, A. Ferrari,et al., “Multi-vendor experimental validation of an open source QoT estimator for optical networks,” J. Light. Technol.36, 3073–3082 (2018)

Show all 39 references
  1. [9]

    GNPy: an open source application for physical layer aware open optical networks,

    A. Ferrari, M. Filer, K. Balasubramanian,et al., “GNPy: an open source application for physical layer aware open optical networks,” J. Opt. Commun. Netw.12, C31–C40 (2020)

  2. [10]

    Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,

    M. Jinno, B. Kozicki, H. Takara,et al., “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag.48, 138–145 (2010)

  3. [11]

    Broadband optical fibre with an attenuation lower than 0.1 decibel per kilometre,

    M. N. Petrovich, E. Numkam Fokoua, Y . Chen,et al., “Broadband optical fibre with an attenuation lower than 0.1 decibel per kilometre,” Nat. Photonics19, 1203–1208 (2025)

  4. [12]

    Low intermodal interference and low loss hollow core fibers,

    P . Li, G. Chen, A. Jia,et al., “Low intermodal interference and low loss hollow core fibers,” inOptical Fiber Communication Conference (OFC) 2026,(2026), p. M2J.1

  5. [13]

    First demonstration of hollow- core-fiber cable for low latency data transmission,

    B. Zhu, B. J. Mangan, T. Kremp,et al., “First demonstration of hollow- core-fiber cable for low latency data transmission,” inOptical Fiber Communication Conference (OFC) 2020, Postdeadline Papers,(2020), p. Th4B.3

  6. [14]

    Transmission of 61 C- band channels over record distance of hollow-core-fiber with L-band interferers,

    A. Nespola, S. Straullu, T. D. Bradley,et al., “Transmission of 61 C- band channels over record distance of hollow-core-fiber with L-band interferers,” J. Light. Technol.39, 813–820 (2021)

  7. [15]

    Beyond terabit/s/ λ nonlinearity-free transmission over the hollow-core fiber,

    Y . Hong, S. Almonacil, H. Mardoyan,et al., “Beyond terabit/s/ λ nonlinearity-free transmission over the hollow-core fiber,” J. Light. Tech- nol.43, 6306–6312 (2025)

  8. [16]

    First demonstration of distributed characterization over 100 km anti-resonant hollow-core fiber in real- time widened C+L-band WDM transmission,

    X. Gao, Q. Zhang, L. Feng,et al., “First demonstration of distributed characterization over 100 km anti-resonant hollow-core fiber in real- time widened C+L-band WDM transmission,” inOptical Fiber Commu- nication Conference (OFC) 2025,(2025), p. Th3F .7

  9. [17]

    Opportunities and challenges for long- distance transmission in hollow-core fibres,

    P . Poggiolini and F . Poletti, “Opportunities and challenges for long- distance transmission in hollow-core fibres,” J. Light. Technol.40, 1605– 1616 (2022)

  10. [18]

    CO2-limited hollow-core fiber links: A capacity-map guide to pre-emphasis and spectral avoidance,

    M. G. Saber and Z. Jiang, “CO2-limited hollow-core fiber links: A capacity-map guide to pre-emphasis and spectral avoidance,” Photon- ics13(2026)

  11. [19]

    Loss in hollow-core optical fibers: mechanisms, scaling rules, and limits,

    E. R. Numkam Fokoua, S. A. Abokhamis Mousavi, G. T. Jasion,et al., “Loss in hollow-core optical fibers: mechanisms, scaling rules, and limits,” Adv. Opt. Photonics15, 1–85 (2023)

  12. [20]

    Modal analy- sis of antiresonant hollow core fibers using S2 imaging,

    A. Van Newkirk, J. E. Antonio-Lopez, J. Anderson,et al., “Modal analy- sis of antiresonant hollow core fibers using S2 imaging,” Opt. Lett.41, 3277–3280 (2016)

  13. [21]

    Ultra-long-haul WDM transmission in a reduced inter-modal interference NANF hollow-core fiber,

    A. Nespola, S. R. Sandoghchi, L. Hooper,et al., “Ultra-long-haul WDM transmission in a reduced inter-modal interference NANF hollow-core fiber,” inOptical Fiber Communication Conference (OFC) 2021,(2021), p. F3B.5

  14. [22]

    Estimation of Kerr nonlinearity in an anti- resonant hollow-core fiber by high-order QAM transmission,

    D. Ge, S. Gao, M. Zuo,et al., “Estimation of Kerr nonlinearity in an anti- resonant hollow-core fiber by high-order QAM transmission,” inOptical Fiber Communication Conference (OFC) 2023,(2023), p. W4D.6

  15. [23]

    Splicing hollow-core fiber with standard glass-core fiber with ultralow back-reflection and low coupling loss,

    B. Shi, C. Zhang, T. Kelly,et al., “Splicing hollow-core fiber with standard glass-core fiber with ultralow back-reflection and low coupling loss,” ACS Photonics11, 3288–3295 (2024)

  16. [24]

    Low loss and broad- band low back-reflection interconnection between a hollow-core and standard single-mode fiber,

    D. Suslov, E. Numkam Fokoua, D. Dousek,et al., “Low loss and broad- band low back-reflection interconnection between a hollow-core and standard single-mode fiber,” Opt. Express30, 37006–37014 (2022)

  17. [25]

    X. Wang, Z. Li, L. Feng,et al., “Simulation investigation of CO 2 absorption-induced performance degradation in hollow-core fiber trans- mission systems and spectrum-optimized digital sub-carrier multiplex- ing design for circumvention,” Opt. Lett.50, 7171–7174 (2025)

  18. [26]

    Hollow-core fiber in direct-detection optical networks: Technology readiness, deployment drivers, and adoption outlook,

    M. G. Saber and Z. Jiang, “Hollow-core fiber in direct-detection optical networks: Technology readiness, deployment drivers, and adoption outlook,” IEEE Netw.early access(2026)

  19. [27]

    Fabrication of long-distance, low-loss and low-gas-absorption hollow core fibre,

    P . Li, G. Chen, A. Jia,et al., “Fabrication of long-distance, low-loss and low-gas-absorption hollow core fibre,” inAsia Communications and Photonics Conference (ACP) 2025,(2025), pp. 1–4

  20. [28]

    Hollow core fiber as a long-term solution for capacity scaling in optical networks,

    G. S. Sticca, M. Ibrahimi, N. Di Cicco,et al., “Hollow core fiber as a long-term solution for capacity scaling in optical networks,” inOptical Fiber Communication Conference (OFC) 2025,(2025), p. M3F .3

  21. [29]

    Hollow-core fiber properties and system-level specifications for next-generation optical transport networks,

    B. Correia and J. Pedro, “Hollow-core fiber properties and system-level specifications for next-generation optical transport networks,” Photonics 13, 71 (2026)

  22. [30]

    Static and dynamic routing, fiber, modulation format, and spectrum allocation in hybrid ULL fiber-SSMF elastic optical networks,

    K. Ouyang, F . Tang, Z. Yuan,et al., “Static and dynamic routing, fiber, modulation format, and spectrum allocation in hybrid ULL fiber-SSMF elastic optical networks,” IEEE Access13, 21652–21664 (2025)

  23. [31]

    ABACUS: An impair- ment aware joint optimal dynamic RMLSA in elastic optical networks,

    M. J. Kiran, V. Chebolu, G. Das, and R. Datta, “ABACUS: An impair- ment aware joint optimal dynamic RMLSA in elastic optical networks,” arXiv preprint arXiv:2404.13308 (2024)

  24. [32]

    Modeling EDFA gain ripple and filter penalties with machine learning for accurate QoT estimation,

    A. Mahajan, K. Christodoulopoulos, R. Martínez,et al., “Modeling EDFA gain ripple and filter penalties with machine learning for accurate QoT estimation,” J. Light. Technol.38, 2616–2629 (2020)

  25. [33]

    Multi-layer protection control for coherent DWDM networks,

    K. Juneja, A. Satbhaiya, and R. Rao, “Multi-layer protection control for coherent DWDM networks,” U.S. Patent 11,310,572 B2 (2022). Assignee: Infinera Corporation. Filed Dec. 18, 2020

  26. [34]

    Characteristics and impacts of CO 2 absorption effects in hollow core fiber (HCF) transmission systems,

    S. Chen, Y . He, L. Dou,et al., “Characteristics and impacts of CO 2 absorption effects in hollow core fiber (HCF) transmission systems,” in European Conference on Optical Communication (ECOC),(2025)

  27. [35]

    Availability design of optical transport networks,

    M. Tornatore, G. Maier, and A. Pattavina, “Availability design of optical transport networks,” IEEE J. on Sel. Areas Commun.23, 1520–1532 (2005)

  28. [36]

    Unavailability analysis of long-haul networks,

    M. To and P . Neusy, “Unavailability analysis of long-haul networks,” IEEE J. on Sel. Areas Commun.12, 100–109 (1994)

  29. [37]

    Parameters and calculation methodologies for reliability and availability of fibre optic systems,

    ITU-T, “Parameters and calculation methodologies for reliability and availability of fibre optic systems,” Recommendation ITU-T G.911, International Telecommunication Union (1997). [Online]. Available: https://www.itu.int/rec/T -REC-G.911-199704-I/en

  30. [38]

    EDFA transient response to channel loss in WDM transmission system,

    A. K. Srivastava, Y . Sun, J. L. Zyskind, and J. W. Sulhoff, “EDFA transient response to channel loss in WDM transmission system,” IEEE Photonics Technol. Lett.9, 386–388 (1997)

  31. [39]

    Output power and SNR swings in cascades of EDFAs for circuit- and packet-switched optical networks,

    L. Tancevski, A. Bononi, and L. A. Rusch, “Output power and SNR swings in cascades of EDFAs for circuit- and packet-switched optical networks,” J. Light. Technol.17, 733–742 (1999)

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.