Quantum Key Management

Abstract

Quantum key distribution (QKD), invented by Bennett & Brassard (1984) based on previous work of Wiesner (1983), has been recognized as a key-technology of the upcoming decades. With various (experimental) quantum networks existing (cf. the reports of Poppe et al. (2008) and Elliott (2004)), questions regarding the efficient construction and management of such networks arise. While much has been achieved in proving security of QKD under various assumptions (trusted devices as proposed by Salvail et al. (2009) vs. non-trustworthy devices as discussed by Elliott (2008b)), and many cryptographic primitives have been transferred to the quantum setting by Buchmann et al. (2004) and Damgard et al. (2004), some questions are still waiting to be answered. With the invention of public-key cryptography, key management has become an issue of major importance. Authentication is equally crucial for QKD-enhanced links, but authenticating keys here is inherently different to the public key setting. Nevertheless, why should quantum cryptography not benefit from the lessons learnt in classic, particularly public-key, cryptography (one of which is the strict principle not to use one key in two different applications)? Elegant ideas for key management and authentication have arisen in public-key cryptography (such as identity-based cryptography invented by Shamir (1985) or certificateless cryptography discussed in Al-Riyami & Paterson (2003)). Are similarly elegant solutions imaginable for the problem of entity authentication in the quantum setting? More importantly, with the one-time pad (OTP) as the encryption of choice, the key demand equals the data transmission demand in terms of size, so an effective management of keys is crucial for a reasonable quality of service of quantum networks. The whole security can be at stake if (quantum-)key generators cannot cope with the flood of information and run empty, thus logically (in terms of secrecy) cutting the link.