Measurement-Device-Independent Quantum Key Distribution (MDI QKD)

QKD is currently the most appealing approach for the secure exchange of a secret key. It removes the vulnerabilities of current classical key distribution methods by providing theoretical security based on the principles of quantum mechanics. However, in real-world implementations, QKD is susceptible to information stealing attacks due to the imperfect nature of the devices utilised. MDI QKD removes the need for a secure detection station which is believed to be the main vulnerability of QKD systems, by transforming it into a transmission station. It however requires the preparation of almost-perfect states and the use of high count rate single photon detection technologies. Previous work carried out within the Engineering Department and Toshiba CRL, has successfully reported methods, via proof-of-principle demonstrations, that satisfy the imposed limitations and produce high key rates. Such methods include the utilisation of gain switched seeded lasers as a way of creating weak coherent pulses that almost perfectly interfere and of self-differencing avalanche photo diodes and superconducting nanowire single photon detectors for high key rates and efficiency at room temperatures.
The project will be focused on the design and optimisation of an MDI QKD system based on previous research, for real life implementation. There is a vast amount of parameters that change in this concept. The aim of this project is to take account of the restrictions of MDI QKD in real situations to progress the current literature, while improving and characterising the optoelectronics used. One of the main issues that needs to be countered is the dramatic and inevitable increase of transmission distances for useful MDI QKD. This deteriorates the conditions under which the system operates and therefore necessitates the development of feedback loops with high stability. Additionally, the synchronisation of the sources and the receiver using a master clock is not viable and alternative methods need to be devised, for example utilising multiplexing. A crucial feature of an implemented MDI QKD is automaticity with real time data analysis and state modulation. Consequently, the development of a code that will drive the suitable optoelectrical components to function and recalibrate the instruments during the run period is mandatory. It is also very significant that true random number generators are used for the selection of bases, bits and any other parameters imposed by the protocols. These generators exploit physical phenomena, usually of quantum nature, that are known to be random and therefore guarantee that their output is completely uncorrelated and unpredictable. Research into true random number generation methods should be carried out to deduce the appropriate method that will optimise the key rates, cost, efficiency and size of the system. Finally, the proof-of-principle demonstrations do not include protection against possible attacks like the Trojan horse, rendering them unfit for commercialisation. Therefore the purpose of the research is not solely to optimise the system under field conditions but additionally to increase its security against hacking attempts.
There are various further ideas that could be researched, depending on the progress and outcomes of the work carried out. Such concepts include the photonic integration of the system or its development into a star shaped network of multiple sources. The collective aim of the project however, remains within the quantum optics and information research area.