Source-linked AI summary
Designing a Quantum Network Protocol
Wojciech Kozlowski, Axel Dahlberg, Stephanie Wehner
TL;DR
Long-distance quantum communication must overcome transmission losses, decoherence, and the inability to clone arbitrary quantum data. The paper designs a quantum data-plane protocol that generates end-to-end entanglement while limiting decoherence, and simulations show it remains functional under decoherence losses and near-term hardware constraints. The protocol is intended as a building block for higher-level quantum network services, but the study leaves complete control-plane design and heterogeneous networks for future work.
Problem
Existing quantum networks cannot forward qubits end-to-end, while transmission losses, decoherence, and no-cloning constrain long-distance communication.
Method
The paper designs a quantum data-plane protocol that coordinates entanglement swapping, reduces memory decoherence, compensates for decoherence losses, and targets sufficient final-pair quality.
Results
The protocol remains functional on extremely resource-limited near-term hardware and delivers its service despite significant decoherence-related losses.
Takeaways & Limitations
The protocol provides a building block for constructing higher-level quantum network services and long-distance end-to-end entanglement.
Takeaways & Limitations
The study focuses on a simplified control plane and homogeneous networks based on a single hardware platform.
Abstract
from arXiv · showhide
The second quantum revolution brings with it the promise of a quantum internet. As the first quantum network hardware prototypes near completion new challenges emerge. A functional network is more than just the physical hardware, yet work on scalable quantum network systems is in its infancy. In this paper we present a quantum network protocol designed to enable end-to-end quantum communication in the face of the new fundamental and technical challenges brought by quantum mechanics. We develop a quantum data plane protocol that enables end-to-end quantum communication and can serve as a building block for more complex services. One of the key challenges in near-term quantum technology is decoherence -- the gradual decay of quantum information -- which imposes extremely stringent limits on storage times. Our protocol is designed to be efficient in the face of short quantum memory lifetimes. We demonstrate this using a simulator for quantum networks and show that the protocol is able to deliver its service even in the face of significant losses due to decoherence. Finally, we conclude by showing that the protocol remains functional on the extremely resource limited hardware that is being developed today underlining the timeliness of this work.
1 INTRODUCTION
Quantum networks aim to provide long-distance quantum communication, but transmission loss, decoherence, and no-cloning constrain current approaches. This paper presents a scalable quantum data-plane protocol for generating end-to-end entanglement and evaluates it under decoherence and near-term hardware constraints.
- Motivation: Current quantum networks cannot forward qubits end-to-end, so longer-distance secure communication remains limited despite established short-distance deployments.Existing longer-distance QKD networks require trusted, physically secure intermediate nodes.
- Motivation: Transmission losses, decoherence, and the no-cloning theorem are the three key challenges for long-distance qubit communication.Typical quantum-network memory lifetimes range from a few microseconds to just over one second.
- Approach: Entanglement swapping connects shorter-range entangled pairs to create long-range entanglement without transmitting qubits across the entire path.Quantum repeaters perform swaps on intermediate qubits, leaving remote qubits entangled.
- Protocol contribution: The proposed protocol coordinates multi-node swapping, reduces memory decoherence, compensates for decoherence losses, and ensures usable final-pair quality.It is designed as a quantum data-plane building block for more complex network services rather than an all-in-one solution.
- Evaluation: The protocol is evaluated in a quantum-network simulator against decoherence and remains functional on extremely resource-limited near-term hardware.These evaluations target the short memory lifetimes and hardware constraints expected in early quantum networks.
2 BACKGROUND AND MOTIVATION
Quantum networks must distribute entanglement despite transmission loss, decoherence, imperfect operations, and the no-cloning theorem. The proposed architecture addresses these constraints by coordinating local quantum operations and classical messages while minimizing qubit storage time.
- Challenges: Quantum networks face transmission losses, decoherence, and the no-cloning theorem when distributing qubits over long distances.Typical quantum-network memory lifetimes range from a few microseconds to just over one second.
- Entanglement distribution: Entanglement swapping connects short-distance entangled pairs into long-distance entanglement without relying on amplification or retransmission.Entangled pairs can be regenerated when lost, enabling a practical alternative to directly transmitting qubits.
- Fidelity and decoherence: Fidelity decreases through imperfect link pairs, swapping, quantum gates, and decoherence during memory storage.Higher link-pair fidelities can improve output quality but come at the cost of reduced generation rates.
- Protocol motivation: The protocol focuses on requesting suitable link fidelities and minimizing idle qubit storage to reduce decoherence.The design treats hardware improvements to gate fidelity as outside the scope of the network protocol.
- Architecture: The network stack coordinates neighboring nodes and local instructions through classical messages, while other protocols handle path computation and related decisions.The stack relies on local operating-system services for operations involved in generating entangled pairs.
3 THE QUANTUM NETWORK LAYER
The quantum network layer delivers entangled pairs with identifiers, state information, fidelity requirements, and timing constraints. It relies on supporting routing, signalling, and link-layer services while the data plane coordinates link generation and entanglement swapping along a path.
- Service requirements: Applications are grouped into “measure directly” and “create and keep” use cases with different tolerance for delivery delays and memory storage.Measure-directly applications consume pairs immediately, whereas create-and-keep applications may store multiple pairs and require closely timed delivery.
- Network-layer service: The network service delivers end-to-end entangled qubits together with an identifier linking the two endpoint qubits.The identifier tracks the link pairs and swaps that produced each long-range pair.
- Network-layer service: The network collects swap announcements so recipients can infer which Bell state each delivered pair occupies.The swapping node learns the state only after the swap completes because the outcome is fundamentally random.
- Quality of service: Applications specify a minimum fidelity threshold and timing requirements such as pair counts, deadlines, rates, or maximum delivery spread.Higher fidelity may require more production time, allowing applications to trade fidelity against rate.
- Network-layer architecture: The quantum data plane coordinates link-level entanglement generation and entanglement swapping along a fixed path, while routing and signalling remain external.Supporting protocols determine paths, establish virtual circuits, manage schedules, and provide required resources.
- Quality of service: The protocol includes quality-of-service mechanisms for checking fidelity thresholds, rejecting infeasible requests, and delaying requests that can be fulfilled later.These mechanisms support quality of service but do not guarantee it independently.
4 QUANTUM NETWORK PROTOCOL
The QNP is a connection-oriented quantum data-plane protocol that establishes virtual circuits, generates and swaps link-level entanglement, and tracks resulting end-to-end pairs. Its design uses lazy tracking, cutoff handling, aggregation, and fidelity properties to support scalable operation under short memory lifetimes.
- Protocol operation: FORWARD messages initiate link-pair generation, swaps occur when adjacent pairs for one circuit are available, and TRACK messages collect swap records at the end-nodes.TRACK messages wait when a corresponding swap has not completed, allowing the end-nodes to infer the final pair information.
- Virtual circuits: The QNP operates over a pre-established virtual circuit, enabling parallel link-pair generation and entanglement swapping along a fixed path.A virtual circuit installs the necessary data-plane state between two end-nodes before protocol operation begins.
- Lazy entanglement tracking: Lazy entanglement tracking records only the final pair state, allowing quantum operations to proceed without waiting for every classical control message.Nodes can discard decohered qubits without separately notifying the rest of the virtual circuit.
- Decoherence handling: The cutoff mechanism discards unswapped qubits that reach a deadline, limiting the use of qubits that may decohere before matching pairs become available.Discard records let later tracking messages notify the originating end-node when the chain is broken.
- Scalability: Aggregation combines requests with the same end-nodes and fidelity threshold, reducing managed circuit state and improving resource sharing at swapping nodes.Shared virtual circuits remove the need to distinguish individual requests when selecting pairs for swaps.
- Fidelity and swapping: Because the fidelity-combination expression is associative, swaps can occur in any order and fidelity can be budgeted differently across heterogeneous links.The protocol therefore retains its swap-order assumption despite variation in link fidelity.
5 EVALUATION
The evaluation examines resource sharing, latency, decoherence robustness, and message-delay effects for the quantum network protocol. Simulations show functional operation across competing circuits and constrained hardware conditions, while also exposing scheduling limitations.
- Evaluation setup: The evaluation uses NetSquid to model quantum hardware effects, including decoherence, propagation delay, fibre losses, and gate operations.The protocol is implemented in Python on top of a link-layer implementation.
- Evaluation setup: The six-node dumbbell topology includes four end-nodes and a bottleneck link, testing entangled-pair flow merging and splitting.All links contain quantum and classical channels.
- Throughput and latency: Latency scales linearly across up to two circuits, but four circuits can trigger a congestion collapse when matching upstream and downstream pairs are unavailable.With only two qubits per link, unswappable pairs can exhaust memory and prevent further pair generation.
- Throughput and latency: A shorter cutoff alleviates four-circuit congestion and improves throughput by discarding unswappable pairs sooner.For one- and two-circuit cases, the shorter cutoff also relaxes link-fidelity requirements by tightening the bound on qubit idle time.
- Decoherence: As memory lifetime decreases, throughput decreases because more qubits are discarded, with higher-fidelity circuits affected more strongly.The cutoff timer outperforms an end-node strategy that uses an oracle to access pair fidelity directly.
- Decoherence: With a memory lifetime of about 1.6 s, message delays have no effect until they approach the cutoff timeout; beyond that threshold, delivered pairs lack sufficient fidelity.The protocol avoids blocking quantum operations while waiting for control messages.
- Resource-limited hardware: The protocol remains functional on resource-limited hardware when routing tables, link fidelities, and cutoff timers are manually tuned.The routing protocol itself does not work well in this environment.
6 DISCUSSION
The paper presents a connection-oriented quantum data-plane protocol for end-to-end entanglement, while identifying missing components needed for complete and heterogeneous quantum-network architectures.
- The protocol delivers end-to-end entanglement across a quantum network and is intended as one component of a complete network architecture.
- Its virtual-circuit design, inspired by MPLS, is intended to support higher-level services such as entanglement distillation, multipath support, and failure recovery.
- The paper focuses on the quantum data plane and uses only a simplified control plane, leaving resource reservation, signalling, and traffic engineering for future work.
- A similar layering-as-optimisation-decomposition approach may offer a systematic way to improve quantum network stack design.
- The study considers homogeneous networks based on a single hardware platform, so protocol performance on hybrid networks remains to be understood.
7 RELATED WORK
Related work includes end-to-end entanglement protocols, repeater-chain constructions, and proposed quantum network stacks, but these approaches differ in scope and functionality.
- Existing end-to-end entanglement proposals differ in their treatment of error correction, decoherence, signalling, and network size.
- Repeater-chain protocols generally target individual linear chains and lack mechanisms for non-linear topologies or merging and splitting flows.
- Prior quantum network-stack proposals provide link-layer designs, functional allocations, or architectural outlines, but some lack concrete network-layer protocols.
8 CONCLUSIONS
The paper presents the Quantum Network Protocol as a data-plane foundation for creating long-distance end-to-end entangled pairs and constructing higher-level quantum-network services.
- The protocol creates long-distance end-to-end entangled pairs, which the paper identifies as a key resource for distributed quantum applications.
- The protocol is intended to serve as a building block for higher-level services that require sophisticated resource management and scheduling.
A ARTIFACTS
The artifact package provides implementation code, simulation data, reproduction instructions, and container support, though compatibility across all platforms is not guaranteed.
- The artifact includes the Quantum Network Protocol implementation in NetSquid and the raw data used to produce the paper’s plots.
- The directory contains simulation source code and data needed to reproduce the reported results.
- README.md documents the artifact contents and setup, while EXPERIMENTS.md explains how to run the experiments and reproduce the data.
- Platform compatibility is not guaranteed, so a Dockerfile is provided for systems supporting Docker containers.
B HARDWARE PARAMETERS
The simulations use NV-centre hardware parameters, with most experiments adopting optimistic beyond-laboratory settings and simplifying qubit resources and noise. A near-future hardware case instead models 25 km fibres and limited communication-qubit availability.
- Hardware platform: The simulations use the nitrogen vacancy centre (NV-centre) platform for quantum repeaters.The paper provides the simulation parameter values in Tables 1 and 2.
- Simulation assumptions: Except for the near-future hardware example, simulations use optimistic parameters beyond current laboratory capabilities.These parameters are compared with currently achievable values in Tables 1 and 2.
- Simulation assumptions: Except in the near-future hardware case, simulations treat all qubits as communication qubits that can participate in link-pair generation.This omits the distinction between communication and memory qubits; the near-future case models only one active link per node and requires more sophisticated scheduling for larger networks.
- Noise and timeout assumptions: The simulations omit communication-qubit-induced memory dephasing and do not calculate cutoff timeouts for idle qubits.The paper treats this noise like decoherence but leaves timeout calculation beyond scope, using a hand-picked timeout in the near-future example.
- Optical fibres: The near-term hardware simulation uses 25 km optical fibres between nodes, while other simulations use 2 m fibres closer to a laboratory scenario.The 25 km case requires frequency conversion to achieve 0.5 dB/km losses.
C.1 Identifiers
The protocol uses identifiers that connect applications, circuits, requests, links, and individual entangled pairs to network nodes and local qubits. These identifiers support routing, link-layer association, and pair tracking across the network.
- Circuit ID: A circuit ID is an opaque handle carried in protocol messages to identify the circuit they concern.The signalling protocol is responsible for circuits.
- Address: An address combines a network-wide node locator with an endpoint identifier on that node.The paper does not specify the exact format of either value.
- Link-pair correlator: A link-pair correlator identifies a generated pair on a particular link and maps it to the corresponding local-memory qubits.It is delivered with the pair by the link layer, and both generating nodes must resolve it locally.
- Link-label: A link-label identifies link-layer requests dedicated to a circuit on each link along its path.Its one-to-one mapping to a circuit ID may differ between links.
- Request ID: A request ID uniquely identifies an application request between two addresses and rejects duplicates.It lets an application reuse an endpoint for multiple requests.
C.2 Messages
The protocol separates messages into request-level control and pair-level tracking, using FORWARD and COMPLETE for requests and TRACK and EXPIRE to manage individual entangled pairs. Node-specific rules coordinate link-pair handling, swapping, expiration, corrections, and delivery.
- Message granularity: The protocol has per-request and per-pair message groups operating at different granularities.Requests concern entangled-pair quantities or rates, while pair-level messages track individual pairs.
- Request-level messages: FORWARD propagates a request from the circuit head-end to the tail-end, initiating or updating link-layer requests and tail-end bookkeeping.Its fields include circuit and request identifiers, endpoint identifiers, request type, pair count, final state, and rate.
- Request-level messages: The head-end calculates the rate used to determine the link-layer request rate, and COMPLETE carries the completed request information.The requested link-rate fraction is based on the circuit’s required end-to-end rate relative to the circuit maximum.
- Pair-level messages: TRACK follows link-pairs and entanglement swaps across the circuit while collecting information needed to infer the end-to-end pair state.It travels in both circuit directions, avoiding a separate receipt acknowledgment.
- Node rules: Protocol rules respond to classical messages and link-layer pairs, with distinct behavior for end-nodes and intermediate swapping nodes.The head-end additionally advances epochs, performs Pauli corrections when required, and sends COMPLETE messages.
- Intermediate-node rules: Intermediate nodes swap an upstream and downstream pair as soon as both are available, while handling message ordering and qubit expiration.LINK, TRACK, and EXPIRE rules are triggered by link-pair arrival, TRACK receipt, and cutoff-timer expiration, respectively.
- Demultiplexing: End-node demultiplexers assign virtual-circuit pairs to requests and cross-check inconsistent symmetric decisions before delivery.NORMAL requests can be reassigned at the tail-end using the head-end decision when the pair has not yet been delivered.
- Algorithm components: The protocol’s helper operations perform Pauli correction, entanglement swapping, and Bell-state combination after swap outcomes.The head-end LINK rule initializes TRACK fields from a newly generated link-pair and sends TRACK downstream.