High-rate, high-dimensional quantum key distribution systems

There is currently a great interest in using high-dimensional (dimension d>2) quantum states for various communication and computational tasks. High-dimensional quantum states provide an efficient and robust means of encoding information, where each photon can encode a maximum of log_2(d) bits of information. One application where this becomes a significant advantage is quantum key distribution (QKD), which is a communication technique that relies on the quantum nature of photonic states to share a classical secret key between two remote users in the presence of a powerful eavesdropper. High-dimensional QKD protocols are believed to overcome some of the practical challenges of the conventional qubit-based (d = 2) protocols, such as the long recovery time of the single-photon detectors, or the low error tolerance to quantum channel noise.

In this thesis, I demonstrate experimentally and theoretically various novel QKD protocols implemented with high-dimensional quantum photonic states, where the information is encoded using the temporal and phase degrees of freedom. One challenging aspect of high-dimensional time-phase QKD protocols is that the measurement of the phase states requires intricate experimental setups, involving time-delay interferometers, fiber Bragg gratings, or a combination of electro-optic modulators and fiber Bragg gratings, among others. Here, I explore two different measurement schemes, one involving a tree of delay line interferometers, and the other using a quantum-controlled technique, where the measurement of the phase states is performed by interfering an incoming quantum state with another locally generated quantum state. Using the interferometric method (quantum-controlled) and a d = 4 (d = 8) encoding scheme, I achieve a secret key rate of 26.2 +/- 2.8 (16.6 +/- 1.0) Mbps at a 4 (3.2) dB channel loss. Overall, the secret key rates achieved in this thesis are a few folds improvement compared to the other state-of-the-art high-rate QKD systems.

Finally, I consider the possibility of an eavesdropper attacking the high-dimensional quantum states using a universal quantum cloning machine, where she uses weak coherent states of different mean photon numbers (decoy-state technique) to estimate the single-photon fidelity. I show that an eavesdropper can estimate the unknown quantum states in the channel with a degraded but optimal cloning fidelity. Specifically, I find that the upper bound of the cloning fidelity decreases from 0.834 +/- 0.003 at d= 2 to 0.639 +/- 0.003 at d = 6, thereby providing evidence for two conclusions. First, the decoy-state technique can be used to extract single-photon contribution from intricate weak coherent states based two-photon experiments. Second, high-dimensional quantum photonic states are more robust compared to the d = 2 quantum states.