What Is Quantum Internet Security? Threats and Defenses in Quantum Communication Infrastructure, Layer by Layer

"It's safe because key distribution is now quantum" is premature. Here's a layer-by-layer breakdown of the OSI model, covering the threats specific to quantum communication infrastructure and how the underlying assumptions change.

The Transport Layer and Quantizing Key Distribution

In the classical internet, TLS performs key exchange at the transport layer (more precisely, spanning OSI layers 4 through 6), relying on computational hardness such as the discrete logarithm problem, as in ECDHE. In the quantum internet, this key exchange itself is replaced with QKD such as BB84, shifting the basis of security from "computational hardness" to "the laws of quantum mechanics." This is a significant change in the sense that key distribution itself remains unaffected even if a future quantum computer breaks public-key cryptography, including ECDHE.

However, it's important to note here that "key distribution has gone quantum" is not the same as "the entire communication has become secure." All QKD provides is a shared bit string used as a key; the actual data payload is still encrypted with a classical symmetric-key cipher such as AES. In other words, what's usually meant by the "quantum internet" is, in practice, a hybrid construction with classical cryptography, where only key distribution has gone quantum.

Security of Quantum Communication Infrastructure by OSI Layer

Physical layer Because the photons themselves carry the information, attacks that exploit flaws in real hardware become a genuine threat. A representative example is the "detector blinding attack," an experimentally demonstrated attack technique that shines strong light on the receiver's single-photon detector to saturate it, forcing it to behave like a classical optical signal, in an attempt to eavesdrop. Also, in implementations that use weak coherent light pulses instead of a true single-photon source, a "photon-number-splitting attack" — where an eavesdropper skims off part of a pulse when it happens to contain multiple photons — is theoretically possible; the decoy-state protocol (mixing in decoy-intensity pulses to detect such splitting attacks) is widely implemented as a countermeasure.
Link layer (quantum repeater networks) A true quantum repeater (a relay that uses only entanglement swapping and never measures or retransmits the key information itself) requires no trust in the relay nodes. However, many QKD networks currently deployed commercially adopt a "trusted node" scheme — where each relay point measures and decrypts the key once, then re-encrypts it for retransmission over the next segment — and in this case every relay node must be trusted, a weakness analogous to that of a classical hop-by-hop VPN.
Transport/session layer QKD guarantees the confidentiality of the key, but authentication is a separate matter. To prevent man-in-the-middle attacks, a classical authentication channel (a pre-shared key, or for the time being a conventional public-key infrastructure) is separately required at the initial stage of communication to guarantee that Alice and Bob are indeed communicating with the correct counterpart. It's easy to overlook that the security of this authentication channel is a hidden assumption underlying the security of the entire QKD system.
Application layer What the application ultimately uses is classical ciphertext, such as AES, encrypted with the key QKD generated. What Toshiba calls a "quantum-safe network" service is, in practice, this same combination of a QKD-derived key with classical cryptography, and the central challenges are down-to-earth implementation issues such as integration with existing network equipment and protocol stacks.
Ad

Summary

  • The quantum internet shifts the basis of security for key exchange from computational hardness to the laws of quantum mechanics
  • However, encryption of the data payload itself is still handled by classical symmetric-key cryptography, making it a hybrid construction
  • At the physical layer, there are threats specific to real hardware, such as the detector blinding attack and the photon-number-splitting attack
  • QKD networks using the trusted-node scheme retain the weakness that every relay node must be trusted
  • QKD guarantees the confidentiality of the key but not authentication, so the security of a separate authentication channel is a prerequisite

Learn How the Quantum Computer Itself Works

The next article explains fault-tolerant quantum computation and the development trends at each company.

Read the Next Article