Distributed Entanglement Distribution Using Multiple Entanglement Sources in WDM-based Quantum Optical Networks

AI summary

Overview

Research area: Quantum communications / quantum optical networks — specifically wavelength-division-multiplexed (WDM) entanglement distribution using spontaneous parametric downconversion (SPDC) sources.

Technical level: Intermediate. Readers should be comfortable with basic photonics concepts (wavelength multiplexing, fiber attenuation, coincidence counting) and with the idea of routing resources through a mesh network.

Scope (one sentence): The paper proposes and simulates a scheme for distributing entangled photon pairs to many node pairs in repeaterless multi-hop mesh optical networks by deploying several narrower-band entangled photon pair sources (EPPSs) at different nodes, rather than one broadband source at a central location.

What This Paper Is About

In a WDM-based quantum network, a single broadband SPDC source emits photons at many correlated wavelength pairs, and each wavelength pair can entangle one pair of nodes. Because one source produces only a limited number of usable wavelength pairs, the number of node pairs that can be entangled is limited, and routing all photons from one central point wastes entangled pairs to fiber and switching loss over long paths.

The paper's goal is to show that placing multiple in-house-characterized EPPSs at different nodes, and letting the network select which source and which wavelength pair serves each entanglement request, uses the finite entanglement resource better and accepts more requests than the centralized single-source approach.

Key Contributions

1. A distributed entanglement distribution approach using multiple SPDC-based EPPSs in WDM-based quantum optical networks, in which each entanglement demand is served by a selected source among several deployed in the network.
2. A candidate node architecture for WDM-based entanglement distribution in multi-hop mesh networks with multiple EPPSs, built from wavelength-selective switches (WSSs) and optical circulators, supporting transmit, receive, and bypass of photons by wavelength.
3. Demand-, loss-, and visibility-aware provisioning: heterogeneous entanglement demands (each with its own required ebit rate and visibility threshold) are served considering estimated degradation of both ebit rate and visibility with fiber length and with the number of hops.
4. A comparative analysis across photon routing schemes (one-photon, OP, versus both-photon, BP distribution) and across centralized versus distributed EPPS placement, in ring, full-mesh, and random partial-mesh topologies generated with the Erdős–Rényi graph model.

Main Findings

• Hybrid OP/BP distribution adds placement flexibility: The paper reports that combining one-photon distribution (source node is one of the two entangling nodes) with both-photon distribution (source node is neither entangling node) gives higher flexibility in placing multiple EPPSs in the network.
• Distributed beats centralized: Distributed entanglement distribution using multiple EPPSs enables better entanglement resource utilization and higher entanglement demand acceptance compared with centralized entanglement distribution from a single broadband source.
• Wavelength pairs differ in quality and rate: Experimental characterization of the in-house three-wavelength-pair source (3WPS) confirms that different wavelength pairs offer different coincidence counts and different entanglement visibilities, motivating demand-aware wavelength-pair assignment rather than arbitrary wavelength selection.
• Usable spectrum is bounded: Six non-overlapping 3WPSs fit in the C-band (each occupying 7 wavelengths at 100 GHz granularity, i.e., 6 × 7 = 42 wavelengths occupying 4.2 THz out of about 4.4 THz), yielding 18 wavelength pairs and thus up to 18 entanglable node pairs; at most four 5WPSs (11 wavelengths each) fit without spectral overlap.
• Visibility is set by loss, not just by source: Because the visibility expression includes accidental coincidence counts divided by the product of the two path transmissions, increasing fiber length (and switching loss, treated as an equivalent fiber length) reduces visibility even when the source visibility is fixed — for the 5WPS the source visibility is roughly constant at about 92%, so the assigned node pair determines the effective visibility.

Methodology in Plain English

The authors first built and measured two entanglement sources in the lab — Sagnac-loop, type-0 SPDC sources based on periodically poled lithium niobate waveguides, one producing three wavelength pairs and one producing five — recording coincidence counts and entanglement visibility for each wavelength pair. Wavelength pairs whose visibility fell below the 71% threshold, and the central "point of degeneracy" wavelength (which cannot be separated deterministically), were excluded.

They then constructed a network model. In a simulator, entanglement requests arrive one at a time, each specifying two nodes, a required ebit rate, and a minimum visibility. For each request, the simulator finds the lowest-loss route between the two nodes, then tries to assign a wavelength pair using a first-fit scheme that also checks whether the pair meets both the rate and visibility requirements after accounting for fiber attenuation, switching loss in optical circulators and WSSs, and added accidental counts. If the request cannot be served from a source at one endpoint, the algorithm tries a source at the other endpoint, and if needed adds a new source at that node (subject to a limit on the number of sources). Two distribution modes are compared: one-photon, where one photon of the pair stays inside the source node and only the other traverses the network, and both-photon, where both photons leave the source node and each takes a separate path, making the pair loss multiplicative.

Several source-placement strategies are compared, including centralized placement and graph-theoretic choices based on betweenness centrality and dominant sets, over ring, full-mesh, and random partial-mesh topologies.

Why This Matters

Impact on research: Prior WDM entanglement distribution demonstrations used a single broadband source in a star topology with a central provider node. This paper reframes source deployment as a network-planning problem — where to put sources, which source serves which request, and which wavelength pair to use — and shows that treating entanglement as a scarce, loss-prone resource to be allocated rather than broadcast is the right framing for mesh networks.

Real-world applications:

• Quantum key distribution between many node pairs sharing one fiber infrastructure, where each pair needs a specified key (ebit) rate and quality.
• Distributed and clustered quantum computing, where entanglement must be delivered at a required rate with a minimum fidelity to interconnected processors.
• Very-long-baseline interferometry and quantum-enhanced navigation, which consume distributed Bell pairs across geographically separated terminals.
• Metro/access quantum network deployments in which a single central source location is impractical or where fiber routes between user sites are long and lossy.

Industry relevance: The scheme reuses standard telecom photonics — 100 GHz grid, C-band (1530–1565 nm), WSSs, optical circulators — meaning it can in principle be planned with the same wavelength-assignment and routing machinery operators already use in classical WDM networks, with added rate and visibility constraints.

Future Directions

• Beyond the C-band: The authors note that only about six 3WPSs (or four 5WPSs) fit in the C-band without spectral overlap, and that multi-band optical networks would allow more sources; quantifying that gain is left open.
• Smarter wavelength-pair assignment: The paper deliberately uses first-fit assignment in all compared cases and explicitly invites more optimal, more complex assignment schemes to improve spectrum utilization.
• Application-level rate accounting: The work stops at entanglement distribution and coincidence detection; it does not include transduction efficiency, memory store/retrieve rates, memory decoherence, buffer size, or classical-communication latency, which would further reduce usable rates.
• Placement optimization at scale: Since betweenness centrality and dominant-set heuristics were used for source placement, an open question is how optimal placement and OP/BP selection strategies scale to larger mesh networks and to non-uniform demand patterns.

Target Audience

Quantum networking researchers and graduate students working on entanglement distribution and quantum network architecture; optical network engineers interested in how WDM planning changes when the "signal" is a finite entangled photon pair; and experimentalists building SPDC sources who want to understand how per-wavelength-pair rate and visibility translate into network-level acceptance of entanglement demands.

Note on quantitative results: the paper content available here is truncated partway through the algorithm description (before the network-level simulation section), so specific simulation values — topologies tested, demand counts, ebit rates, acceptance ratios, and visibility after multi-hop routing — are not reported in the text provided and are therefore not stated above.

Authors’ abstract

Quantum network implementations using single spontaneous parametric downconversion (SPDC)-based broadband entangled photon pair source (EPPS) have been reported recently. Here, leveraging the wavelength-correlation between entangled photon pairs, the traditional wavelength division multiplexing (WDM) method is utilized to route photons based on their wavelengths. From single EPPS, entangled photon pairs are distributed in a centralized way to different node pairs in the network. However, the number of nodes pairs that can be entangled in a network is limited by the number of entangled wavelength pairs that an EPPS can generate. To entangle a higher number of node pairs in a network, multiple EPPSs can be employed. In this work, we present a WDM-based entanglement distribution approach using multiple EPPSs in multi-hop repeaterless mesh optical networks. We experimentally characterize two EPPSs developed in-house and consider multiple such EPPSs in the network to perform network-level simulations. We consider heterogeneous entanglement demands requiring different entanglement bit (ebit) rates and entanglement visibility. For each entanglement demand, EPPS placement/selection, wavelength-pair assignment, and photon pair routing are performed considering the degradation in both ebit rate and visibility with fiber length and hops in the network. Two main findings of this study include: (i) a hybrid approach of one-photon (OP) and both-photon (BP) entanglement distribution provides higher flexibility in multi-EPPS placement and (ii) distributed entanglement distribution using multiple EPPSs enables better entanglement resource utilization and higher entanglement demand acceptance as compared to centralized entanglement distribution.