Quantum Information Networks

• Quantum Information Networks are entanglement-based infrastructures that generate, route, store, and process quantum states using teleportation and entanglement swapping.
• They integrate quantum and conventional networking components like repeaters and base stations to overcome physical impairments such as decoherence and limited memory lifetimes.
• Performance metrics including fidelity, throughput, and probability of success, alongside physics-informed routing strategies, optimize QIN operations for both local and space-based deployments.

Searching arXiv for the provided QIN papers and closely related works to ground the article. Quantum Information Networks (QIN) are networks in which quantum information is generated, distributed, routed, stored, processed, and consumed across interconnected quantum devices. In the networking literature, a QIN is described as “a specific network concept which relies on generating, distributing and routing entanglement of bipartite quantum systems between many users,” with teleportation and entanglement swapping as the core operational mechanisms and entanglement as the fundamental network resource (Paccard et al., 1 Aug 2025). In the quantum-internet literature, the same object is framed as a network of systems that can “generate, store, process, send, and receive quantum information,” with end nodes, quantum channels, and quantum repeaters as the core components (Kubala et al., 5 Feb 2026). In both formulations, QIN differs from classical internetworking because quantum states are fragile, non-clonable, probabilistic, and perishable, so end-to-end service is often realized by creating and managing entanglement rather than forwarding a payload directly (Chehimi et al., 2022).

1. Scope, definition, and conceptual boundaries

In the communications and satellite literature, QIN is broader than quantum key distribution. It is intended to connect quantum computers, sensors, memories, and end users through entanglement distribution, entanglement swapping, and teleportation, supporting secure quantum communications, distributed quantum computing, and distributed quantum sensing rather than only key establishment (Paccard et al., 1 Aug 2025). The same literature repeatedly stresses that QIN is not a simple extension of packet networking to qubits: quantum states cannot in general be copied or amplified, quantum repeaters are not amplifier-like devices, and the classical network remains necessary for coordination, heralding, signaling, and control (Abelem et al., 2023).

A recurring misconception is that QIN can be analyzed by transplanting classical networking abstractions into a quantum setting. Chehimi and Saad explicitly argue that this is not viable, because many networking-oriented analyses assume idealized quantum hardware, such as perfect entanglement generation, lossless channels, perfect repeaters, or infinite-lifetime quantum memories (Chehimi et al., 2022). Their central thesis is that practical quantum communication networks are a physics-constrained systems problem whose architecture, protocols, performance metrics, and optimization methods must be grounded in the behavior of qubits, photons, memories, and entanglement operations.

The term also has a broader, nonstandard usage. In “Information Processing by Networks of Quantum Decision Makers,” a multi-agent society of decision makers exchanging information through a Hilbert-space formalism is described as “a kind of a quantum information network,” with node dynamics governed by the decomposition

$p(\pi_n)=f(\pi_n)+q(\pi_n),$

where $f$ is a utility factor and $q$ is an attraction factor (Yukalov et al., 2017). This broader usage is conceptually important, but the dominant usage in quantum networking research refers to physical entanglement-centric infrastructures rather than quantum-inspired decision systems.

2. Physical principles and elementary operations

The basic information unit is the qubit, written as

$\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$

with $|\alpha|^2+|\beta|^2=1$, and practical network states are often represented by density operators

$\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$

(Abelem et al., 2023). Bell pairs such as $(\ket{00}+\ket{11})/\sqrt{2}$ are the canonical bipartite network resource, while multipartite states such as GHZ states appear in secret sharing, Byzantine agreement, and multiparty protocols (Abelem et al., 2023).

Teleportation and entanglement swapping are the central QIN primitives. Teleportation consumes one Bell pair plus classical communication to transfer a quantum state, while entanglement swapping concatenates short entangled links into longer ones. Because arbitrary unknown quantum states cannot be copied, classical repeater logic does not carry over. Quantum repeaters therefore partition long links into shorter segments and use Bell state measurements rather than amplification (Chehimi et al., 2022). The same no-cloning constraint also rules out classical-style redundancy, broadcast replication, and amplifier-like repeater behavior (Paccard et al., 1 Aug 2025).

Practical operation is dominated by physical impairments. The core list emphasized in physics-informed network design includes decoherence, finite quantum-memory lifetime, limited memory capacity, probabilistic entanglement generation, frequency-conversion loss, channel loss and noise, entanglement-swapping constraints, and fidelity degradation during transmission, storage, and imperfect operations (Chehimi et al., 2022). For lossy channels, the explicit success model given is

$P_s=e^{-\beta d},$

where $P_s$ is the probability that a photon successfully traverses a channel segment, $\beta$ is the attenuation coefficient, and $f$0 is the transmission distance (Chehimi et al., 2022). State quality is tracked by fidelity; the standard density-operator expression explicitly alluded to is

$f$1

with $f$2 (Chehimi et al., 2022).

These constraints make timing and storage first-class network variables. Bell-state measurements require relevant photons or qubits to be available with sufficient fidelity at the same time, but the waiting induced by synchronization and heralding further degrades the stored states (Chehimi et al., 2022). The effect is architectural: in QIN, the physical layer is not merely below the network layer; it shapes feasibility, service quality, and scheduling state.

3. Architectures, layers, and hybrid control

The component model of QIN typically includes quantum end nodes, single-photon sources, emitters, photonic quantum channels, matter-qubit memories, quantum repeaters, entanglement-distribution mechanisms, and network-control entities such as a quantum base station or coordinating access point (Chehimi et al., 2022). End nodes consume entanglement or quantum states for communication, sensing, or distributed computing; memories buffer states while waiting for other segments, heralding messages, purification rounds, or scheduling decisions; and repeaters act as stochastic entanglement-processing nodes rather than signal boosters (Chehimi et al., 2022).

Layered views have emerged, but they remain cross-coupled. One influential protocol stack cited in the quantum-internet literature includes a Physical Entanglement layer, Entanglement Control, Error Management, Quantum State Propagation, and Applications (Abelem et al., 2023). Physical Entanglement creates Bell pairs over adjacent links, Entanglement Control coordinates single-hop entanglement attempts and qubit selection, Error Management handles purification and quality improvement, Quantum State Propagation decides when to purify and when to swap, and the Application layer requests end-to-end entanglement or consumes it directly (Abelem et al., 2023).

A central architectural result is that QIN is intrinsically hybrid. “Lessons Learned on the Interface between Quantum and Conventional Networking” proposes a quantum-conventional network harness in which the control plane carries control and management traffic, while the data plane handles the conventional and quantum data communications (Alshowkan et al., 2021). In that deployment, practical quantum networking required dedicated fiber for photonic transport together with encrypted conventional network connections for control traffic, timestamps, and experiment coordination (Alshowkan et al., 2021). The same systems view appears in fielded service platforms: the reconfigurable Quantum Internet Service Provider developed at the Center for Quantum Networks abstracts hardware resources through a modular, open-source framework called Quagent and exposes them through a Platform-as-a-Service model (Yang et al., 2023).

The distinction between physical and logical topology is already important in local-area systems. In the Oak Ridge quantum-conventional network, a star physical topology combined with spectral switching yielded a richer logical transport graph (Alshowkan et al., 2021). In the University of Arizona QISP testbed, the physical network is a two-level star centered at the ECE building, but the logical layer forms a 13-node complete graph because centralized entangled-photon sources, SNSPDs, and switches can provision entanglement between arbitrary terminal nodes (Yang et al., 2023).

4. Metrics, quality measures, and cross-layer optimization

Physics-informed QIN research replaces classical packet-network metrics with hardware-aware figures of merit. Chehimi and Saad identify throughput, fidelity, Quality of Matter Qubits (QoMQ), delay, and Probability of Success (PoS) as central performance measures (Chehimi et al., 2022). In this formulation, throughput is not raw bit rate but the number of entangled qubits successfully generated, preserved, and delivered; fidelity is an end-to-end service-quality metric; QoMQ is measured through absorption efficiency into the matter qubit, coherence time, and extraction efficiency back from the matter qubit; delay includes waiting for probabilistic photon generation, graph-state preparation, heralded Bell-state-measurement outcomes, purification attempts, and propagation; and PoS attaches success probabilities to channels and operations (Chehimi et al., 2022).

These metrics induce genuinely quantum control variables. The paper points to the number of photons in logically encoded states, the number of single-photon sources used for multiplexing, quantum-memory usage, storage-time scheduling, and emitter–matter qubit coupling as design variables (Chehimi et al., 2022). A representative trade-off is drawn from NV centers: stronger nuclear-spin coupling to the electron spin enables faster two-qubit gates for storage and retrieval, reducing delay, but it also induces more unwanted interactions and thus lower fidelity; weaker coupling improves fidelity at the cost of slower operations (Chehimi et al., 2022). Because the environment is stochastic and success probabilities are hardware-driven, reinforcement learning is suggested as a promising control method for scheduling policies that balance storage time against fidelity loss (Chehimi et al., 2022).

Routing quality has also been recast in explicitly state-aware terms. “Trace-Distance based End-to-End Entanglement Fidelity with Information Preservation in Quantum Networks” models a quantum network as

$f$3

with edge capacities $f$4, finite node memories $f$5, and a centralized controller that collects fidelity and trace-distance information (Kumar et al., 2024). Its core metric for information preservation is the trace distance

$f$6

while path selection uses closeness centrality

$f$7

The resulting Trace-Distance based Path Purification algorithm selects candidate paths via closeness centrality, diagnoses the edge with the path’s worst trace-distance condition, applies entanglement pumping there, and then proceeds to entanglement swapping, using a fidelity threshold of $f$8 in the protocol flow (Kumar et al., 2024). The same paper also states that some equations are malformed or physically nonstandard, and explicitly characterizes the update $f$9 as unusual physically (Kumar et al., 2024). That limitation is instructive: state-aware routing is increasingly central, but the formalism is still unsettled.

5. Routing, coding, and distributed state representations

Several recent directions treat routing and coding themselves as quantum resources. In “Quantum networks with coherent routing of information through multiple nodes,” a message carrier is routed in a coherent superposition of different paths rather than being forced onto a single route (Kristjánsson et al., 2022). The decisive interference term is governed by the vacuum interference operator

$q$0

and Theorem 1 shows that asymptotic nonzero classical capacity under coherent routing is possible precisely when $q$1 has singular value $q$2 (Kristjánsson et al., 2022). In the idealized Z-channel example, coherent routing yields an asymptotic classical capacity

$q$3

while fixed-path transmission would have vanishing capacity (Kristjánsson et al., 2022). For finite chains with path dephasing, the interference term is reduced by

$q$4

but lower-bound capacity gains remain for moderate length and small path noise (Kristjánsson et al., 2022).

A related but distinct proposal is channel aggregation. “Aggregating Quantum Networks” encodes a logical state into a quantum Reed–Solomon code

$q$5

and distributes different physical qudits over different channels, so that the network itself becomes part of a distributed erasure-correcting substrate (Piparo et al., 2020). The target success threshold is

$q$6

and the central result is that transmission can succeed even when no individual route alone has sufficient resources or sufficient quality (Piparo et al., 2020). Combined with spatial-temporal single-photon multiplexing, the paper reports a significant drop in network resources required to transmit the quantum signal (Piparo et al., 2020).

Distributed encoding and storage also admit fully network-theoretic formulations. “Distributed Encoding and Decoding of Quantum Information over Networks” studies tree-topology networks and proves that exact spreading of quantum information is achievable if and only if, for every tree edge $q$7,

$q$8

where $q$9 is the reduced state of the encoded reference state across the corresponding cut (Yamasaki et al., 2018). The same work shows that concentrating quantum information can require strictly less entanglement than spreading it, even though encoding and decoding are inverse tasks (Yamasaki et al., 2018). “Delocalized information in quantum networks” pushes the idea further by storing logical information nonlocally across a region or the whole network and transporting it using local measurements on individual nodes only, with explicit constructions based on stabilizer codes, Dicke states, and correlation-space or matrix-product-state resources (Miguel-Ramiro et al., 2019).

Higher-layer task orientation has also entered QIN. “Quantum Semantic Communications for Resource-Efficient Quantum Networking” proposes a pipeline in which classical data are preprocessed, embedded in a high-dimensional Hilbert space by a quantum feature map, clustered by quantum $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$0-means into semantic concepts, and then mapped into a lower-dimensional semantic latent space for transport (Chehimi et al., 2022). Its communication objective is constrained by both communication fidelity and semantic fidelity, and in the modeled setting it reports approximately $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$1 reduction in quantum communication resources at $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$2 and about $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$3 reduction at $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$4 relative to semantic-agnostic baselines, while achieving higher quantum semantic fidelity (Chehimi et al., 2022). The corresponding implication is that QIN may optimize for semantic utility rather than raw state delivery.

6. Space-based deployment, service infrastructures, and major applications

Global-scale QIN has made satellites a central architectural question. “Entanglement Management in Space-Based Quantum Information Networks” and “The Role of the Satellite in Quantum Information Networks” both argue that satellites are indispensable for global connectivity because fiber loss grows exponentially with distance and quantum states cannot be amplified noiselessly (Paccard et al., 1 Aug 2025, Paccard et al., 1 Aug 2025). A benchmark cited in the space-based architecture paper is stark: with a perfect $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$5 GHz single-photon source, ideal detectors, and fiber attenuation of $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$6, over $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$7 one would detect only about $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$8 photon per century on average (Paccard et al., 1 Aug 2025). The same paper divides architectures into those without quantum optical inter-satellite links and those with QOISL, and distinguishes source satellites, repeater satellites with Bell-state-measurement devices and memories, and hybrid satellites (Paccard et al., 1 Aug 2025). Its conservative conclusion is that the only architecture realistic in the near future is a downlink satellite carrying an entangled-photon source (Paccard et al., 1 Aug 2025).

The quantitative comparison between ground and space paths is sharper in “The Role of the Satellite in Quantum Information Networks.” There, the space path becomes “significantly useful” for connecting users $\ket{\psi}=\alpha\ket{0}+\beta\ket{1},$9 away; at about $|\alpha|^2+|\beta|^2=1$0, the satellite path is better by about $|\alpha|^2+|\beta|^2=1$1 orders of magnitude than the optimized ground path with six swapping nodes; and at about $|\alpha|^2+|\beta|^2=1$2, it is better by about $|\alpha|^2+|\beta|^2=1$3 orders of magnitude (Paccard et al., 1 Aug 2025). The reported average fidelities are $|\alpha|^2+|\beta|^2=1$4 for the space path in short, medium, and long scenarios, while ground fidelities degrade from $|\alpha|^2+|\beta|^2=1$5 to $|\alpha|^2+|\beta|^2=1$6 to $|\alpha|^2+|\beta|^2=1$7 across the same scenarios (Paccard et al., 1 Aug 2025).

Service-oriented local deployments are developing in parallel. Oak Ridge National Laboratory’s quantum-conventional network connected three quantum laboratories over dedicated fiber and conventional network connections, showing how a control plane, a hybrid data plane, secure tunnels, synchronized timing, and centralized timestamp analysis are necessary even for local entanglement-distribution experiments (Alshowkan et al., 2021). At the University of Arizona, the Center for Quantum Networks built a reconfigurable QISP over a campus fiber network spanning 5 buildings and 13 terminal laboratory nodes, with centralized entangled-photon sources, an eight-channel SNSPD array, optical switches, and a web-based orchestration framework; the resulting physical star topology yields a logical 13-node complete graph and supports multi-channel entanglement distribution, routing, and concurrent services for multiple users (Yang et al., 2023).

Applications already extend far beyond secure key exchange. The quantum-internet literature lists distributed quantum computing, blind quantum computing, computing on encrypted data, improved sensing, secure private-bid auctions, clock synchronization, telescope interferometry, quantum voting, Byzantine agreement, secret sharing, and cloud access to remote quantum computers as network-enabled tasks (Abelem et al., 2023). The HEP roadmap emphasizes quantum-assisted telescopes, quantum clock networks, magnetometer arrays, searches for axionlike and scalar particles, and remote quantum processing of detector outputs such as gamma spectra (Derevianko et al., 2022). In quantum-assisted telescope proposals, photon indistinguishability is constrained by

$|\alpha|^2+|\beta|^2=1$8

and at $|\alpha|^2+|\beta|^2=1$9 this is written as

$\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$0

while clock networks are explicitly contrasted with classical ones by the scaling $\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$1 versus $\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$2 for $\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$3 clocks (Derevianko et al., 2022).

7. Open problems and evolving interpretations

The field’s dominant open problem is still practical viability. Physics-informed networking identifies unresolved issues in signal design, scheduling and resource management under short-lived states, distributed quantum computing over QCNs, interoperability across device platforms, and the integration of photonic communication qubits with trapped ions, superconducting qubits, annealers, and other computing substrates (Chehimi et al., 2022). The quantum-internet survey adds hardware transparency, reactive control under limited qubit lifetimes, scalable routing over virtual quantum links, link-layer multiplexing, and the absence of a settled notion of a “quantum packet” as open systems problems rather than purely physical ones (Abelem et al., 2023). The 2026 review places similar weight on standardization, transduction, repeaters, synchronization, compiler support, and heterogeneous interfaces (Kubala et al., 5 Feb 2026).

Space-based QIN adds its own unresolved constraints. Downlink is preferred to uplink because uplink is more severely degraded by atmospheric turbulence through point-ahead and anisoplanatic angles; Bell-state measurements can require near-perfect time synchronization on the order of $\rho \equiv \sum_i p_i \ket{\psi_i}\bra{\psi_i}$4; and the same survey states that embarking a space quantum memory is unrealistic at the moment (Paccard et al., 1 Aug 2025). These are not peripheral engineering details. They directly determine which satellite roles are credible in the near term and which remain long-term targets.

Finally, QIN remains an evolving umbrella term. In physical networking, it denotes entanglement-centric infrastructures whose resources are generated, stored, swapped, routed, and consumed across hybrid classical-quantum systems (Paccard et al., 1 Aug 2025). In quantum-inspired information processing, it can also denote distributed Hilbert-space dynamics over multi-agent nodes (Yukalov et al., 2017). The coexistence of these usages reflects the breadth of the field, but the research frontier in networking is now clear: practical QIN demands physics-informed architectures, state-aware routing and scheduling, interoperable control software, deployable memories and repeaters, and service abstractions that can support communication, sensing, and distributed computation on the same network substrate.