Short answer: Device-independent quantum key distribution (DI-QKD) is a form of QKD whose security is certified by observed correlations that violate a Bell inequality, so it does not need a trusted model of the devices. Review articles describe it as the strongest form of implementation security, and also as very challenging in practice, with only proof-of-principle demonstrations so far. This is a research area, not a product category.
The idea
A 2023 review in npj Quantum Information says DI-QKD provides what it calls the gold standard for secure key exchange. It says that, besides allowing information-theoretic security based on quantum mechanics, it relaxes the need to physically model the devices, which rules out many quantum hacking threats to which non-DI QKD systems are vulnerable. A 2023 review in Quantum describes it as often seen as the ultimate key exchange protocol in terms of security, since it can be performed securely with uncharacterised black-box devices. A review in IET Quantum Communication says it relies solely on observed nonlocal correlations, certified through Bell inequality violations, removing assumptions about the internal workings of the measurement devices.
The problem it addresses
The IET review says that, despite its theoretical security promise, practical QKD systems are vulnerable to serious attacks, including side-channel attacks and detector loopholes, and assume a trusted device characterisation. The npj paper calls DI-QKD the ultimate solution to the problem of implementation security because it does not require characterising the internal functioning of any device.
How it works, conceptually
The npj review says DI-QKD builds on the historical Ekert 91 protocol, in which a central untrusted source distributes entangled photon pairs between two parties, Alice and Bob, each with a measurement unit, and the violation of a Bell inequality signals the security of the quantum channel. By performing local measurements, the parties can certify monogamous correlations between their outcomes on the basis of their input-and-output statistics alone. The IET review says it examines protocol designs including CHSH-based schemes and non-local game frameworks, along with their security proofs.
Why it is hard
The npj review says DI-QKD relies on the loophole-free violation of a Bell inequality, which requires high quality entanglement to be distributed between distant parties and close to perfect quantum measurements, which it calls hardly achievable with current technology. It adds that recent theoretical and experimental efforts have led to proof-of-principle DI-QKD implementations. The Quantum review likewise says implementing DI-QKD is technically challenging, and that there have been recent proof-of-principle demonstrations.
The IET review lists open problems: noise tolerance, generation rates and integration with quantum networks. It also covers recent experimental implementations, source architectures, detection technologies and finite-key analyses.
What this does not tell you
This article summarises the abstracts and introductions of three review articles. It does not give key rates or distances for any experiment, and it does not claim that DI-QKD is available as a product.
FAQ
What does device-independent mean?
Security is certified from observed Bell inequality violations, without a trusted model of the internal workings of the devices.
Why is it called the gold standard?
The npj review uses that phrase because it gives information-theoretic security while ruling out many quantum hacking threats.
Has it been demonstrated?
Reviews say proof-of-principle demonstrations exist. They also say it is very challenging with current technology.
What does it need?
Loophole-free Bell inequality violation, which needs high quality entanglement between distant parties and close to perfect measurements, per the npj review.
What are the open problems?
The IET review names noise tolerance, generation rates and integration with quantum networks.
All field notes