Thirty years after Bennett and Brassard have proposed the first Quantum Key Distribution (QKD) protocol, QKD - which allows to share secret keys at distant via quantum communication - has certainly developped as the main "practical" application of quantum information, with expected applicaton in network security. We will recall briefly the steps of this history, and will describe the level of development of QKD today as well as the challenges and opportunities it is facing.
We will then focus on one particular technology, namely Continuous Variable QKD, initially proposed in the group of Philippe Grangier, at IOTA, in collaboration with Thales, ans subsequently developed at Telecom ParisTech and the spin-off SeQureNet, co-founded by Romain Alléaume. CVQKD, relying on coherent detection, is an appealing QKD technology as experimental systems can be build entirely with telecom-grade components, while CVQKD performance have progressed notably in the past few years, with maximum distance now in the 100 km range. We will explain how CVQKD works and two recent results obtained by our group :
- the demonstration of DWDM compatiblity of CVQKD (http://arxiv.org/abs/1412.1403)
- the question of implementation security and recent results on side-channels attacks on CVQKD.
We will then discuss a question that may find resonance with the expertise and research performed at LPL in metrology and synchronization of clocks over optical netwoks, namely the question of phase locking in CVQKD.
The question is related to the optimum design of an experimental CV-QKD system working in a regime where an independent laser (phase-locked with the laser at Alice) is used at the reception side (Bob).
We will review the results obtained in the two articles
and present our work, namely original protocols and simulations together with some open questions related to the best choices for a future experimental implementation.
Finally if time permits, we will evoke a new proposal for quantum cryptography (for secure message passing and key distribution) based on a hybrid security model involving time-release classical encryption and noisy quantum memories.