4.1.6 Implementation and Security

Printer-friendly versionSend by emailPDF version

Physical approaches and perspectives

QKD is synonymous with security– the security is based on the laws of quantum physics. However, as with any security technologies there are always potential weaknesses due to imperfections in the implementation. In the case of QKD, this has given rise to the field of “quantum hacking” as researchers attempt to find potential side-channels, implementation weaknesses, which mean that the device is no longer described by its, quite often abstract, security model. There would appear to be two ways that one can deal with this, either build better devices that have no implementation flaws, or, define the security in a way that is independent of the device and its implementation. The applied effort is focused on the first possibility with companies and (ethical) hackers working closely. Nonetheless, there is a clear need to bridge the gap between these abstract security models and practical implementations with minimal assumptions. Recently researchers have revisited an idea that has been around since the first proposal for entanglement-based QKD [1]. This approach has been labelled Device Independent QKD (DI-QKD) as it treats the devices as black boxes– the security is dependent solely on calculating a few input-output probabilities to calculate a relatively simple inequality– a Bell inequality. If the inequality is violated, the system is secure, independent of the internal workings of the device. Concepts such as device independent security and more recently, self-testing systems, provide a new paradigm, not only for security, but for characterising complex, distributed, quantum networks.
European groups working in this field include: A. Acin (ICFO, ES), S. Massar and S. Pironio (Brussels,  BE), R. Renner (Zurich, CH), J. Skaar (NTNU, Norway), R. T. Thew and N. Gisin (Geneva, CH).

State of the art

Dealing with implementation issues at an applied level requires close collaboration between quantum hackers and those building, and even selling, the QKD systems and technologies. This approach has been well demonstrated for recent detection attacks [2].  DI-QKD, on the other hand, is a relatively new concept and its experimental application requires unprecedented performance of the systems and component technologies. Nonetheless, a couple of recent papers have started to bring this into the realms of experimental feasibility [3, 4]. Central to this was the concept of heralded photon amplifiers [5], which have also been realised experimentally in the visible [6, 7] and more recently, telecom regimes[8]. Self-Testing is another related concept where the effort is to minimise assumptions and to help better characterise quantum systems and technologies. To date this has primarily been a theoretical effort for the moment  [9-11]. The adaptation and demonstration of DI-QKD will also be important for future secure networks. In a further extension of this idea, heralded photon amplifiers have been proposed in a recent quantum repeater protocol [12] that is not only one of the most efficient but it also hints at the potential for DI scenarios across quantum networks. 


Both quantum hacking and device independent security have similar goals, but approach the task from opposite directions: Both have the aim of minimising the assumptions involved in secure quantum communication systems and to bridge the gap between the theoretical proofs and the security of the final implementation. European theory groups have been a driving force in this area, especially for the later, although experimental initiatives have already started in several European groups, as well as in Singapore, Canada and Australia. Some of the key challenges are:

  • Security proofs for QKD systems that are optimised to cope with a wide range of experimental parameters including finite key lengths;
  • Experimental implementations that minimise side-channels and information leakage;
  • Experimental demonstration of self-testing concepts;
  • Lab demonstration of device-indpendent QKD. 

Key references
[1] A. Ekert, Phys. Rev. Lett. 67, 661 (1991)
[2] L. Lydersen et al., Nature Photonics 4, 686 (2010)
[3] A. Acin et al., Phys. Rev. Lett. 98, 230501 (2007)
[4] N. Gisin, S. Pironio and N. Sangouard, Phys. Rev. Lett. 105, 070501 (2010)
[5] T. Ralph and A. Lund, Quantum Communication Measurement and Computing Proceedings of 9th International Conference, Ed. A.Lvovsky, 155 (AIP, New York 2009) - arXiv:0809.0326v1 (2009)
[6] G. Y. Xiang et al., Nature Photonics 4, 316 (2010)
[7] F. Ferreyrol et al. Phys. Rev. Lett. 104, 123603 (2010)
[8] C. I. Osorio et al., Phys. Rev. A 86, 023815 (2012)
[9] D. Mayers and A. Yao, Quantum Inform. Comput. 4 (2004)
[10] M. McKague, T. H. Yang, V. Scarani, arXiv:1203.2976v1 (2012)
[11] C. C. W. Lim et al., arXiv:1208.0023 (2012)
[12] J. Minar, H. de Riedmatten and N. Sangouard, Phys. Rev. A 85, 032313 (2012)