🌐 Beyond bits: the fundamental shift
Today's internet carries information using classical bits — sequences of zeros and ones traveling through optical fibers. The quantum internet will use qubits: quantum bits that can exist in a superposition of two states simultaneously, encoded in the polarization of photons or the spin of electrons. The fundamental difference, however, is not digital capacity — it is the physical impossibility of interception. In a quantum network, any attempt to measure a quantum state irreversibly alters it, immediately revealing the presence of a third party.
This property derives from two foundational principles: the no-cloning theorem, which forbids creating an exact copy of an unknown quantum state, and the measurement-disturbance principle, which means that measurement inevitably disturbs the measured system. Together, they form the basis of quantum cryptography — a field that began with a lecture by two scientists at a hotel in India, forty years ago.
🔐 BB84 and E91: the protocols that started everything
In 1984, Charles Bennett (IBM) and Gilles Brassard (Université de Montréal) presented the BB84 protocol in Bangalore — the first method of Quantum Key Distribution (QKD). The logic was simple but radical: Alice sends individual photons in random polarization bases (rectilinear or diagonal), Bob measures them in random bases as well, and then they compare which bases matched over a public classical channel. The bits where bases match become the key. If an eavesdropper Eve attempted to measure the photons in transit, she would introduce errors — Alice and Bob need only 72 bits of comparison to detect interception with 99.9999999% probability.
Seven years later, Artur Ekert (Oxford) proposed an entirely different approach: the E91 protocol relies on pairs of entangled photons. Instead of sending individual photons, a pair of photons in a Bell state is created — if Alice measures vertical polarization, Bob will automatically measure vertical as well, regardless of distance. Security is based on violation of Bell inequalities: if a third party intervened, they would introduce local realism to the system, the statistical test S would deviate from the maximum value 2√2, and the interception would become visible.
⚙️ Anatomy of a quantum network
The structure of a quantum network architecturally resembles the classical internet, but every component operates radically differently. At the endpoints are quantum processors (end nodes) — small quantum systems capable of generating, storing, and measuring qubits. They can be implemented using NV-centers in diamond (which operate at room temperature), ion traps, or cavity quantum electrodynamics (Cavity QED) systems. The transmission lines are existing telecommunications optical fibers — no new cabling is needed — but photons travel as single-photon pulses with a mean photon number below 1.
The greatest technical challenge is distance. Classical networks use amplifiers every few kilometers, but in the quantum world this is impossible: due to the no-cloning theorem, you cannot copy a qubit to amplify it. The solution is the quantum repeater, which works through entanglement swapping. Two pairs of entangled qubits are created — one between sender and repeater, and one between repeater and receiver. A Bell measurement at the repeater “teleports” the entanglement to the two endpoint qubits, doubling the distance without physical transfer of a qubit.
🏗️ Who is building quantum internet today?
The first serious attempt was the DARPA network (2003-2007): 10 nodes in Massachusetts, in collaboration with BBN Technologies, Harvard, and Boston University. In Europe, the SECOQC network (2003-2008) connected 6 locations in Vienna using 200 km of optical fiber. In Geneva, SwissQuantum (2009-2011) linked CERN with the University — its goal was transitioning from laboratory to production environment. In Tokyo (2010), seven organizations (NEC, Mitsubishi, NTT, Toshiba, Id Quantique) achieved for the first time one-time-pad encryption at practical rates, sufficient for secure video conferencing.
China took the baton in the most ambitious way. In August 2016, Pan Jianwei's team launched the Micius satellite (QUESS mission), which achieved entanglement distribution over a distance of 1,203 km between two ground stations. In September 2017, the Beijing-Shanghai trunk line was inaugurated: a 2,000 km QKD network with 32 relay nodes that immediately processed banking transactions. In 2021, researchers combined 700+ optical fibers with two satellite QKD links, creating an integrated 4,600 km network — the largest quantum communication network on Earth.
In the US, the IQNET program (Caltech + AT&T + Fermilab, 2017) achieved qubit teleportation over 44 km of optical fiber in December 2020. In Bristol (2020), an 8-user city-scale network operated on existing fiber infrastructure without trusted nodes. In Delft (2022), data teleportation was extended for the first time to three locations instead of two. In February 2025, Oxford researchers demonstrated distributed quantum computation between two ion traps separated by 2 meters, executing Grover's algorithm with a 71% success rate.
⚠️ Obstacles and criticism
Skeptics are not lacking — and they are not just theoreticians. The US National Security Agency (NSA) identifies five fundamental problems: (1) QKD provides only a partial solution, as it does not authenticate the sender — classical cryptography is needed in combination, (2) it requires specialized equipment and dedicated fibers, (3) trusted nodes increase cost and insider threat risk, (4) actual security depends on hardware — something much harder to certify than software, (5) post-quantum cryptography may offer equivalent security without special equipment.
In practice, the longest QKD distance in optical fiber is 833.8 km — but at a very low transmission rate. Practical records stand at 307 km (University of Geneva / Corning, 2015) and 380 km without trusted nodes (IIT Delhi, 2023). In 2024, South African and Chinese scientists achieved atmospheric QKD at a record 12,900 km via microsatellite, transferring over one million quantum-secure bits in a single orbit.
🔮 The future: when and how?
Singapore inaugurated NQSN+ (National Quantum-Safe Network Plus) in 2023, targeting a nationwide quantum-safe network by 2030. Europe is developing EuroQCI (European Quantum Communication Infrastructure). China is planning to expand the trunk line to cities along the Yangtze River. In the US, the IETF created the QIRG (Quantum Internet Research Group) and the ITU established a Focus Group on quantum information in networks (FG-QIT4N) — standards are being prepared.
Technologically, the breakthrough will come when quantum repeaters become practical. Today, networks rely on trusted nodes — nodes that must be physically secure. A true quantum repeater would enable end-to-end security without trust in intermediaries. Platforms such as NV-centers in diamond, ion traps, and quantum modems (Max Planck Garching, 2020) show the path — but scaling remains a challenge.
Quantum internet does not mean faster internet. It means physically impossible to intercept internet. Security is not based on mathematical problems that might someday be solved — it is based on laws of nature.
The first banking transaction via QKD took place in Vienna in 2004. The first use in national elections was in Switzerland in 2007, for transmitting ballot results in Geneva. Commercial use is already here — companies like ID Quantique (Geneva), Toshiba, and QNu Labs (India) sell QKD systems. But a full quantum internet, with end-to-end quantum security at global scale, remains 10-15 years away. The physics already works. The engineering is what needs to catch up.