A theoretical study of subcarrier-wave quantum key distribution system integration with an optical transport network utilizing dense wavelength division multiplexing

AI summary

Overview

Research area: Quantum communications, specifically quantum key distribution (QKD) integrated with classical dense wavelength division multiplexing (DWDM) optical transport networks.

Technical level: Intermediate. The paper is theoretical and numerical, and it assumes some familiarity with QKD security analysis and fiber nonlinear optics.

Scope: A theoretical and numerical study of how spontaneous Raman scattering, four-wave mixing, and filter crosstalk degrade subcarrier-wave QKD when it shares fiber with classical DWDM channels, comparing C-band and 1310 nm O-band quantum-channel placement.

What This Paper Is About

QKD normally works best on dedicated dark fiber, but deploying separate fiber is expensive and impractical as traffic grows. A more attractive option is to multiplex weak quantum signals alongside strong classical DWDM channels in the same fiber, but those classical channels generate Raman scattering, four-wave mixing, and filter leakage that can overwhelm the quantum signal. This paper models those noise sources for a subcarrier-wave QKD (SCW QKD) system and calculates how far secure keys can be generated under different DWDM configurations and quantum-channel wavelengths.

Key Contributions

1. Extends earlier SCW-QKD integration analysis by including four-wave mixing and linear channel crosstalk alongside spontaneous Raman scattering, which the authors’ earlier work did not account for.
2. Builds a noise-to-detection-probability model that converts Raman, four-wave-mixing, and crosstalk power into photon-detection probabilities, then into quantum bit error rate and secure key generation rate.
3. Evaluates C-band DWDM scenarios in which the quantum channel replaces one channel in a uniform grid, including 8-channel grids with 100 GHz and 50 GHz spacing and a 40-channel grid.
4. Compares C-band operation with O-band operation at 1310 nm, including dark-fiber baselines, and derives the filter extinction required for O-band performance to approach dark-fiber performance.

Main Findings

• Raman scattering dominates in the C-band: For most DWDM configurations with the quantum channel inside a uniform C-band grid, spontaneous Raman scattering is the dominant noise source for SCW QKD.
• Four-wave mixing depends on spacing, Raman noise on channel count: Four-wave mixing noise is mostly governed by the DWDM channel spacing, while spontaneous Raman scattering mainly depends on the number of classical channels.
• Four-wave mixing matters only in specific regimes: Its impact becomes significant for configurations with a small number of channels and low channel spacing. At the maximum achievable distance in the studied 8-channel, 100 GHz case, four-wave mixing was negligible compared with Raman scattering and even channel crosstalk.
• Maximum distance shrinks with denser, larger grids: For an 8-channel DWDM grid with 100 GHz spacing, the maximum achievable distance was 32 km. Reducing the spacing to 50 GHz with the same number of channels reduced it to 26 km. Increasing to 40 channels reduced it to 8 km.
• Longer distances require narrow filtering and fewer classical channels: The results indicate that achieving distances above 50 km requires a smaller filter bandwidth and integration with a low number of classical channels.
• O-band placement strongly suppresses Raman noise: Placing the quantum channel in the O-band reduces spontaneous Raman scattering noise by approximately 4000 times.
• O-band fiber loss is higher: Optical losses at 1310 nm are almost twice as large as at 1550 nm, and in dark fiber the maximum achievable distance at 1310 nm is approximately twice as small as in the C-band.
• O-band still wins with many classical channels: Despite the higher attenuation, the SCW-QKD system performs much better at 1310 nm than in the C-band in the presence of 40 classical channels with 100 GHz spacing.
• Crosstalk limits the O-band case: In the O-band configuration with 40 classical channels, the largest noise contribution comes from linear channel crosstalk.
• Filter extinction above 110 dB is needed: To achieve performance close to dark fiber at 1310 nm, the extinction of the quantum-channel filtering system should be higher than 110 dB.
• Specific O-band parameters used: The comparison used fiber attenuation of ξ = 0.35 dB/km at 1310 nm and receiver sensitivity of Rx = −23 dBm for the information channels.

Methodology in Plain English

The authors treat the problem analytically and numerically rather than experimentally. They start with standard expressions for forward and backward spontaneous Raman scattering generated by classical channels, using the ITU frequency grid definition ν_n (THz) = 191.6 + N ν_spacing. Classical channel output power is set from receiver sensitivity and insertion loss, P_out (dBm) = Rx (dBm) + IL (dB), so that the classical DWDM system still meets its bit-error-rate requirements.

They then add two other noise mechanisms. Four-wave mixing is calculated from the input powers of triplets of classical channels, the fiber nonlinearity coefficient, the degeneracy factor, the attenuation, and a phase-matching efficiency that depends on chromatic dispersion and channel frequency separations. Linear channel crosstalk is calculated from the finite isolation of the filter that separates the quantum channel from the classical channels, P_LCXT (dBm) = P_out (dBm) − ISOL (dB).

Each noise power is converted into a photon-detection probability using the detector quantum efficiency, detector gating time, and receiver insertion loss. These probabilities are inserted into an approximate QBER expression that also includes dark counts, phase misalignment, modulation depth, and filter efficiency. The secure key generation rate is then estimated using an asymptotic Devetak–Winter-type bound with a Shannon entropy term for error correction and a Holevo term for a collective beam-splitting attack. The analysis assumes infinitely long keys, independent and identically distributed states, and collective attacks; Eve is assumed not to be affected by the channel noise. The “maximum achievable distance” is extracted from the calculated key rate versus fiber length.

The numerical study covers C-band grids in which one DWDM channel is replaced by the quantum channel, including 8 channels at 100 GHz spacing, 8 channels at 50 GHz spacing, and 40 channels. It also compares four cases at 1532.7 nm and 1310 nm, each with dark fiber and with DWDM. The fixed system parameters include ξ = 0.18 dB/km, μ0 = 3.93, m = 0.316, Dλ = 16.0 ps/(nm × km), dDλ/dλ = 0.09 ps/(nm² × km), p_dark = 4 × 10⁻⁶, Δφ = 5°, Δλ = 15 GHz, Ω = 4.8 GHz, η_D = 0.1, ϑ = 10⁻³, IL = 8 dB, v_S = 100 MHz, γ = 0.78 W⁻¹ km⁻¹, and Δt = 1 ns.

Why This Matters

The paper gives network designers a quantitative reason to place quantum channels in the O-band rather than packing them into the C-band DWDM grid, even though 1310 nm has higher fiber attenuation. It also shows that filter isolation, not only Raman noise, becomes the limiting factor in that preferred configuration. For operators, this informs whether QKD can be added to existing DWDM infrastructure without dedicated dark fiber.

Real-world applications:

• Telecom metro and backbone networks: Adding QKD to existing DWDM transport links without deploying new fiber.
• Quantum-safe data-center interconnects: Sharing fiber between classical high-capacity traffic and quantum key distribution.
• Government, financial, and critical-infrastructure links: Upgrading deployed optical networks with quantum-derived keys.
• Plug-and-play QKD upgrades: Replacing a single DWDM channel’s SFP modules with QKD transmitter and receiver modules, as the paper discusses for the C-band replacement case.

Industry relevance: The work addresses the cost and deployment barrier that has kept many QKD systems on dark fiber. It also highlights practical component requirements, especially filter extinction above 110 dB, which directly affects optical filter and transceiver design for coexisting quantum and classical channels.

Future Directions

• Experimental validation: The paper is entirely theoretical and numerical; laboratory or field trials with SCW QKD and classical DWDM traffic would test the predicted distance limits and filter requirements.
• Extension to continuous-variable QKD: The authors state that the analysis can be adapted to continuous-variable QKD schemes by expressing the calculated noise as excess noise in the covariance matrix.
• Optimized channel allocation: The paper cites an existing optimized allocation scheme that accounts for both four-wave mixing and Raman noise; further optimization across larger and denser DWDM grids is a natural next step.
• Finite-key and more advanced security analysis: The current key-rate calculation uses the asymptotic infinite-key case and a collective beam-splitting attack; finite-key effects and other attack models remain open.
• Filter and isolation engineering: Achieving extinction higher than 110 dB while maintaining narrow bandwidth is a practical open problem for O-band integration.

Target Audience

This paper is most useful to QKD researchers, optical transport network engineers, and network planners evaluating whether quantum and classical channels can share deployed DWDM fiber. It also suits graduate students and researchers working on fiber-noise modeling, dense wavelength multiplexing, or quantum-safe optical infrastructure. Readers without a background in QKD or fiber nonlinear optics will find the detailed noise and security-rate modeling challenging.

Authors’ abstract

In this paper we study the performance of subcarrier-wave quantum key distribution (SCW QKD) in the presence of classical channels of optical transport network utilizing dense wavelength division multiplexing (DWDM). We consider the impact of spontaneous Raman scattering noise as well as the four-wave mixing and channel isolation efficiency. We calculate quantum bit error rate as well as the secure key generation rate of SCW-QKD protocol for different parameters of DWDM system and quantum channel allocations. Our calculations show, that quantum channel allocation at the wavelength of 1310 nm is preferable to allocation at C-band due to lower nonlinear noise, despite fiber's larger attenuation at this wavelength.