Quantum Network

A quantum network connects quantum devices using links that carry quantum information rather than classical bits. The defining capability is not raw bandwidth but the ability to distribute entanglement between distant nodes, which unlocks applications impossible on any classical network.

Quantum vs classical networks

Classical networks move information by copying it: a router reads a packet, stores it, and forwards a copy. This works because classical bits can be copied freely. Quantum networks cannot operate this way. The no-cloning theorem forbids copying an unknown quantum state, and measuring a qubit to “read” it destroys the quantum information it carries.

Instead, quantum channels transmit qubits directly, preserving their quantum state from sender to receiver. The fundamental resource a quantum network produces is not data throughput but shared entangled pairs between nodes that may be far apart. Those pairs are then consumed by applications: key distribution, teleportation, or coordination between distributed quantum processors.

Node types

A quantum network contains several categories of nodes:

End nodes are the sources and destinations of quantum communication. They may be quantum computers, quantum sensors, or simple devices capable only of generating and measuring individual qubits. An end node needs enough quantum capability to perform its application, but not necessarily a full-scale quantum processor.

Repeater nodes sit along the links between end nodes. Their job is to extend entanglement across distances that exceed what a single photon link can span. They store one end of an entangled pair in quantum memory while waiting for the other link to succeed, then perform entanglement swapping to join the links together.

Quantum routers are a more advanced concept: nodes that can switch quantum channels dynamically and route entanglement toward its destination. True quantum routing remains largely theoretical, as it requires quantum memory, fast switching, and the ability to handle multiple simultaneous entangled paths.

Entanglement distribution as the key primitive

The central operation in a quantum network is generating a shared entangled pair between two nodes. Everything else builds on top of it. Quantum teleportation consumes one entangled pair to transmit a qubit. Quantum key distribution uses entangled pairs (in the Ekert protocol) to generate a secret key. Distributed quantum computing links separate processors by using teleportation to pass qubits between them.

Distributing entanglement over long distances requires quantum repeaters because direct photon transmission through fiber is limited to roughly 100-200 km before loss becomes prohibitive.

Quantum memory requirements

A quantum network without memory is possible but severely constrained. Without memory, two nodes can only establish entanglement if both ends of a photon link succeed simultaneously, which becomes increasingly unlikely as links get longer and losses mount.

Quantum memory allows a node to hold one end of an entangled pair while waiting for the other link to complete. This changes the scaling from a product of probabilities (very bad) to something much more favorable. The required storage time grows with the number of links being combined and the distance involved. Current quantum memories can hold states for times ranging from microseconds to several seconds depending on the technology, with the best demonstrations in rare-earth crystals and trapped-ion systems.

Current state: city-scale QKD networks

As of 2026, deployed quantum networks are primarily quantum key distribution systems. China operates the largest: a network connecting Beijing, Shanghai, and other cities, spanning over 2,000 km using a combination of fiber links and the Micius satellite as a relay. European projects including the OpenQKD initiative have deployed QKD testbeds across multiple cities.

These existing networks mostly use trusted nodes: points where quantum signals are measured and re-transmitted classically, which means the relay operator must be trusted. They are not full quantum networks in the sense of end-to-end coherent quantum links, but they demonstrate the operational and engineering challenges involved.

The quantum internet roadmap

Stephanie Wehner and colleagues proposed a six-stage roadmap for the quantum internet, progressing from simple trusted-relay networks toward a fully connected quantum internet:

Stage 1: Trusted-node networks (classical relay, quantum link only at each hop). Stage 2: Prepare-and-measure networks (end-to-end qubit transmission without entanglement). Stage 3: Entanglement distribution networks (shared entangled pairs between any two nodes). Stage 4: Quantum memory networks (stored entanglement, enabling more complex protocols). Stage 5: Fault-tolerant few-qubit networks (limited error correction at nodes). Stage 6: Quantum computing networks (full-scale quantum processors connected by quantum channels).

Current deployments are at stage 1. Academic demonstrations have reached stage 3 over short distances.

Applications

Quantum key distribution uses quantum links to establish encryption keys with information-theoretic security, meaning security that does not depend on the hardness of any mathematical problem.

Distributed quantum computing connects separate quantum processors into a larger effective machine, allowing computations too large for any single device.

Enhanced quantum sensing uses entanglement between spatially separated sensors to achieve measurement precision beyond what individual sensors can reach. Telescope arrays linked by quantum channels could achieve resolution equivalent to an Earth-sized telescope.

Blind quantum computing lets a client delegate computation to a remote quantum server without the server learning what was computed, using protocols that depend on quantum communication between client and server.

See also

• Quantum Internet
• Quantum Repeater
• Quantum Key Distribution
• Entanglement
• Quantum Teleportation