Encrypted data transmission is becoming increasingly more important as information security is vital to modern communication networks. Quantum Key Distribution (QKD) is a promising method based on the quantum properties of light to generate and distribute unconditionally secure keys for use in classical data encryption. Significant progress has been achieved in the performance of QKD point-to-point transmission over a fibre link between two users. The transmission distance has exceeded several hundred kilometres of optical fibre in recent years, and the secure bit rate achievable has reached megabits per second, making QKD applicable for metro networks. To realize quantum encrypted data transmission over metro networks, quantum keys need to be regularly distributed and shared between multiple end users. Optical switching has been shown to be a promising technique for cost-effective QKD networking, enabling the dynamic reconfiguration of transmission paths with low insertion loss. In this thesis, the performance of optically switched multi-user QKD systems are studied using a mathematical model in terms of transmission distance and secure key rates. The crosstalk and loss limitations are first investigated theoretically and then experimentally. The experiment and simulation both show that negligible system penalties are observed with crosstalk of -20 dB or below. A practical quantum-safe metro network solution is then reported, integrating optically-switched QKD systems with high speed reconfigurability to protect classical network traffic. Quantum signals are routed by rapid optical switches between any two endpoints or network nodes via reconfigurable connections. Proof-of-concept experiments with commercial QKD systems are conducted. Secure keys are continuously shared between virtualised Alice-Bob pairs over effective transmission distances of 30 km, 31.7 km, 33.1 km and 44.6 km. The quantum bit error rates (QBER) for the four paths are proportional to the channel losses with values between 2.6% and 4.1%. Optimising the reconciliation and clock distribution architecture is predicted to result in an estimated maximum system reconfiguration time of 20 s, far shorter than previously demonstrated. In addition, Continuous Variable (CV) QKD has attracted much research interest in recent years, due to its compatibility with standard telecommunication techniques and relatively low cost in practical implementation. A wide band balanced homodyne detection system built from modified off-the-shelf components is experimentally demonstrated. Practical limits and benefits for high speed CVQKD key transmission are demonstrated based on an analysis of noise performance. The feasibility of an optically switched CV-QKD is also experimentally demonstrated using two virtualised Alice-Bob pairs for the first time. This work represents significant advances towards the deployment of CVQKD in a practical quantum-safe metro network. A method of using the classical equalization technique for Inter-symbol-interference mitigation in CVQKD detection is also presented and investigated. This will encourage further research to explore the applications of classical communication tools in quantum communications.

<p>The thesis presents the basics of Quantum Key Distribution, a survey of the present techniques, a look at the possible future, and finally a comparison to the alternative technique of using public key or manual distribution of keys.</p><p>Techniques to integrate QKD with the existing telecom fiber infrastructure have been studied, and so has the EU-funded project SECOQC.</p><p>Last the security and efficiency of QKD have been examined, with focus on what level of security that is required, existing security solutions have been used as a comparison.</p>

This thesis describes the deployment of a free-space quantum key distribution system across the University of Waterloo campus. The quantum key distribution system has the ability to provide unconditionally secure communication between two parties: Alice and Bob. The system exploits the quantum mechanical property of entanglement in order to generate a key. Security is then guaranteed by the No-Cloning theorem and the laws of quantum mechanics which prevent a quantum system from being measured without disturbing it. Polarization-entangled photon pairs are created using the non-linear optical process of type-II spontaneous parametric down-conversion. A free-space link of approximately $\mathrm{580~m}$ is used to distribute one-half of the pairs to Alice at a distant location, while the other half of the pairs are locally detected by Bob. The details of the detection apparatus necessary to measure the polarization of the photons and the software used to process the measurement data according to the BBM92 protocol are described. An experimental violation of the CHSH inequality (a derivative of the original Bell inequality) is demonstrated to show that polarization-entangled photon pairs are in fact being distributed to the two parties. Finally, the full BBM92 protocol is performed using the entangled photon pairs to generate a secure key and transmit an encrypted message between Alice and Bob. Currently, the system can only be operated at night because background light saturates the detectors during the day; however, future work will focus on making daylight operation feasible.

Many protocols and experiments in quantum information science are described in terms of simple measurements on qubits. However, in an experimental implementation, the exact description of the measurement is usually more complicated. If there is a claim made from the results of an experiment by using the simplified measurement description, then do the claims still hold when the more realistic description is taken into account? We present a "squashing" model that decomposes the realistic measurement description into first a map, followed by a simplified measurement. The squashing model then provides a connection between a realistic measurement and an ideal measurement. If the squashing model exists for a given measurement, then all claims made about a measurement using the simplified description also apply to the complicated one. We give necessary and sufficient conditions to determine when this model exists. We show how it can be applied to quantum key distribution, entanglement verification, and other quantum communication protocols. We also consider several examples of detectors commonly used in quantum communication to determine if they have squashing models.

This thesis describes the deployment of a free-space quantum key distribution system across the University of Waterloo campus. The quantum key distribution system has the ability to provide unconditionally secure communication between two parties: Alice and Bob. The system exploits the quantum mechanical property of entanglement in order to generate a key. Security is then guaranteed by the No-Cloning theorem and the laws of quantum mechanics which prevent a quantum system from being measured without disturbing it. Polarization-entangled photon pairs are created using the non-linear optical process of type-II spontaneous parametric down-conversion. A free-space link of approximately $\mathrm{580~m}$ is used to distribute one-half of the pairs to Alice at a distant location, while the other half of the pairs are locally detected by Bob. The details of the detection apparatus necessary to measure the polarization of the photons and the software used to process the measurement data according to the BBM92 protocol are described. An experimental violation of the CHSH inequality (a derivative of the original Bell inequality) is demonstrated to show that polarization-entangled photon pairs are in fact being distributed to the two parties. Finally, the full BBM92 protocol is performed using the entangled photon pairs to generate a secure key and transmit an encrypted message between Alice and Bob. Currently, the system can only be operated at night because background light saturates the detectors during the day; however, future work will focus on making daylight operation feasible.

Many protocols and experiments in quantum information science are described in terms of simple measurements on qubits. However, in an experimental implementation, the exact description of the measurement is usually more complicated. If there is a claim made from the results of an experiment by using the simplified measurement description, then do the claims still hold when the more realistic description is taken into account? We present a "squashing" model that decomposes the realistic measurement description into first a map, followed by a simplified measurement. The squashing model then provides a connection between a realistic measurement and an ideal measurement. If the squashing model exists for a given measurement, then all claims made about a measurement using the simplified description also apply to the complicated one. We give necessary and sufficient conditions to determine when this model exists. We show how it can be applied to quantum key distribution, entanglement verification, and other quantum communication protocols. We also consider several examples of detectors commonly used in quantum communication to determine if they have squashing models.

Quantum key-distribution (QKD) is a scheme for establishing shared secret key between remote parties. In such a scheme, quantum preparation and measurement devices (sources and detectors) are used. In existing theoretical treatments of QKD, the device models used do not capture all the imperfections which might occur in realistic devices. This creates a gap between the practical implementations and theoretical descriptions of QKD. In the present work, we contribute in bridging this gap by three methods: 1) Advancing the study of squashing models of measurement devices, 2) Devising an alternative to squashing models using statistical estimation in optical QKD, and 3) Modifying the security proof formalism of QKD to account for imperfect devices.

Quantum key distribution (QKD) allows two remote parties to generate secret keys for cryptographic purposes. Its security has been proven with some assumptions. However, practical realizations may not comply with all the assumptions, leading to various attacks. Founded on the observation that almost all attacks are on the detection part, measurement-device-independent QKD (MDI-QKD) was proposed to remove all such attacks. This thesis presents an implementation of the protocol. In our implementation, key bit information was encoded in the polarization states of weak coherent pulses at 1542 nm wavelength in optical fibers, and decoy state techniques were employed. We ensured stable polarization preparation and alignment and developed a QKD system over 10 km of standard Telecom fibers at 500 KHz repetition rate. Our work demonstrates the practicality of MDI-QKD protocol of removing all attacks, existing and yet to be discovered, on the detection part of a QKD system.

Quantum key distribution (QKD) allows two remote parties to generate secret keys for cryptographic purposes. Its security has been proven with some assumptions. However, practical realizations may not comply with all the assumptions, leading to various attacks. Founded on the observation that almost all attacks are on the detection part, measurement-device-independent QKD (MDI-QKD) was proposed to remove all such attacks. This thesis presents an implementation of the protocol. In our implementation, key bit information was encoded in the polarization states of weak coherent pulses at 1542 nm wavelength in optical fibers, and decoy state techniques were employed. We ensured stable polarization preparation and alignment and developed a QKD system over 10 km of standard Telecom fibers at 500 KHz repetition rate. Our work demonstrates the practicality of MDI-QKD protocol of removing all attacks, existing and yet to be discovered, on the detection part of a QKD system.

A major field of research at the current time is that of implementing Quantum Key Distribution over large distances using satellites. If this protocol works with this technology, it will have huge implications on future information security. In order for a satellite to implement this idea, there are many aspects that must be taken into account. One of the big issues that comes up for this type of system is that of propagating light through the turbulent atmosphere and its effects on the acquisition, pointing and tracking system. The projects studied in this thesis study some of the effects of the atmosphere on certain detectors, try to develop pointing schemes for better accuracy as well as develop knowledge in free space propagation of other single photon experiments. In the first experiment, I study the spatial correlations of the daughter photons created in spontaneous parametric down conversion. I look at the effect of altering the pump beam on the positions of the down converted photons and see if the pump can be manipulated in a way to control the directions of the daughter photons. I begin to utilize a deformable mirror and Shack-Hartmann wavefront sensor which are generally used in adaptive optics, but we plan to use them to alter the pump beam in the spontaneous parametric down conversion process to analyze the correlations between the pump and down converted photons. The second experiment investigates the effects of laser scintillation on the performance of a possible tracking device that could be implemented on a satellite. This quad sensor tracks the position of a beam and a system will be developed to move the sensor to keep the beam in the center where there is a hole for the quantum single photons to stream through. In order to create the effects of scintillation, a turbulence simulator box was built and characterized. This box combines wind turbulence with a heat gradient to mimic atmospheric turbulence on a small scale. Finally, my contributions to a large scale, long distance free space quantum optics experiment are explained and the overall goal of the experiment is discussed. This experiment exposed me to actual free space transmission issues as well as many fundamental techniques for performing long distance optics experiments. In this experiment there was no correction for atmospheric turbulence, but in the future, techniques could be implemented which might increase the efficiencies of the free space links.