S'il fallait résumer le sujet de cette thèse en une expression, cela pourrait être quelque chose comme: phénomènes de grande dimension (mais néanmoins finie) en théorie quantique de l'information. Cela étant dit, essayons toutefois de développer brièvement. La physique quantique a inéluctablement affaire à des objets de grande dimension. Partant de cette observation, il y a, en gros, deux stratégies qui peuvent être adoptées: ou bien essayer de ramener leur étude à celle de situations de plus petite dimension, ou bien essayer de comprendre quels sont les comportements universels précisément susceptibles d'émerger dans ce régime. Nous ne donnons ici notre préférence à aucune de ces deux attitudes, mais au contraire oscillons constamment entre l'une et l'autre. Notre but dans la première partie de ce manuscrit (Chapitres 5 et 6) est de réduire autant que possible la complexité de certains processus quantiques, tout en préservant, évidemment, leurs caractéristiques essentielles. Les deux types de processus auxquels nous nous intéressons sont les canaux quantiques et les mesures quantiques. Dans les deux cas, la complexité d'une transformation est mesurée par le nombre d'opérateurs nécessaires pour décrire son action, tandis que la proximité entre la transformation d'origine et son approximation est définie par le fait que, quel que soit l'état d'entrée, les deux états de sortie doivent être proches l'un de l'autre. Nous proposons des solutions universelles (basées sur des constructions aléatoires) à ces problèmes de compression de canaux quantiques et d'amenuisement de mesures quantiques, et nous prouvons leur optimalité. La deuxième partie de ce manuscrit (Chapitres 7, 8 et 9) est, au contraire, spécifiquement dédiée à l'analyse de systèmes quantiques de grande dimension et certains de leurs traits typiques. L'accent est mis sur les systèmes multi-partites et leurs propriétés ayant un lien avec l'intrication. Les principaux résultats auxquels nous aboutissons peuvent se résumer de la façon suivante: lorsque les dimensions des espaces sous-jacents augmentent, il est générique pour les états quantiques multi-partites d'être à peine distinguables par des observateurs locaux, et il est générique pour les relaxations de la notion de séparabilité d'en être des approximations très grossières. Sur le plan technique, ces assertions sont établies grâce à des estimations moyennes de suprema de processus gaussiens, combinées avec le phénomène de concentration de la mesure. Dans la troisième partie de ce manuscrit (Chapitres 10 et 11), nous revenons pour finir à notre état d'esprit de réduction de dimensionnalité. Cette fois pourtant, la stratégie est plutôt: pour chaque situation donnée, tenter d'utiliser au maximum les symétries qui lui sont inhérentes afin d'obtenir une simplification qui lui soit propre. En reliant de manière quantitative symétrie par permutation et indépendance, nous nous retrouvons en mesure de montrer le comportement multiplicatif de plusieurs quantités apparaissant en théorie quantique de l'information (fonctions de support d'ensembles d'états, probabilités de succès dans des jeux multi-joueurs non locaux etc.). L'outil principal que nous développons dans cette optique est un résultat de type de Finetti particulièrement malléable / If a one-phrase summary of the subject of this thesis were required, it would be something like: miscellaneous large (but finite) dimensional phenomena in quantum information theory. That said, it could nonetheless be helpful to briefly elaborate. Starting from the observation that quantum physics unavoidably has to deal with high dimensional objects, basically two routes can be taken: either try and reduce their study to that of lower dimensional ones, or try and understand what kind of universal properties might precisely emerge in this regime. We actually do not choose which of these two attitudes to follow here, and rather oscillate between one and the other. In the first part of this manuscript (Chapters 5 and 6), our aim is to reduce as much as possible the complexity of certain quantum processes, while of course still preserving their essential characteristics. The two types of processes we are interested in are quantum channels and quantum measurements. In both cases, complexity of a transformation is measured by the number of operators needed to describe its action, and proximity of the approximating transformation towards the original one is defined in terms of closeness between the two outputs, whatever the input. We propose universal ways of achieving our quantum channel compression and quantum measurement sparsification goals (based on random constructions) and prove their optimality. Oppositely, the second part of this manuscript (Chapters 7, 8 and 9) is specifically dedicated to the analysis of high dimensional quantum systems and some of their typical features. Stress is put on multipartite systems and on entanglement-related properties of theirs. We essentially establish the following: as the dimensions of the underlying spaces grow, being barely distinguishable by local observers is a generic trait of multipartite quantum states, and being very rough approximations of separability itself is a generic trait of separability relaxations. On the technical side, these statements stem mainly from average estimates for suprema of Gaussian processes, combined with the concentration of measure phenomenon. In the third part of this manuscript (Chapters 10 and 11), we eventually come back to a more dimensionality reduction state of mind. This time though, the strategy is to make use of the symmetries inherent to each particular situation we are looking at in order to derive a problem-dependent simplification. By quantitatively relating permutation symmetry and independence, we are able to show the multiplicative behavior of several quantities showing up in quantum information theory (such as support functions of sets of states, winning probabilities in multi-player non-local games etc.). The main tool we develop for that purpose is an adaptable de Finetti type result

We investigate the geometric phase (GP) acquired by the states of a spin-1/2 nucleus which is subject to a static magnetic field. This nucleus as the carrier system of GP, is taken as coupled to a dissipative environment, so that it evolves non-unitarily. We study the effects of different characteristics of different environments on GP as nucleus evolves in time. We showed that magnetic field strength is the primary physical parameter that determines the stability of GP / its stability decreases as the magnetic field strength increases. (By decrease in stability what we mean is the increase in the time rate of change of GP.) We showed that this decrease can be very rapid, and so it could be impossible to make use of it as a quantum logic gate in quantum information theory (QIT). To see if these behaviors differ in different environments, we analyze the same system for a fixed temperature environment which is under the influence of an electromagnetic field in a squeezed state. We find that the general dependence of GP on magnetic field does not change, but this time the effects are smoother. Namely, increase in magnetic field decreases the stability of GP also for in this environment / but this decrease is slower in comparison with the former case, and furthermore it occurs gradually. As a second problem we examine the entanglement of two atoms, which can be used as a two-qubit system in QIT. The entanglement is induced by an external quantum system. Both two-level atoms are coupled to a third two-level system by dipole-dipole interaction. The two atoms are assumed to be in ordinary vacuum and the third system is taken as influenced by a certain environment. We examined different types of environments. We show that the steady-state bipartite entanglement can be achieved in case the environment is a strongly fluctuating, that is a squeezed-vacuum, while it is not possible for a thermalized environment.

Made available in DSpace on 2016-06-02T20:16:49Z (GMT). No. of bitstreams: 1 4363.pdf: 3900088 bytes, checksum: fd6c9860d749b37b41d8610825d05459 (MD5) Previous issue date: 2012-03-16 / Financiadora de Estudos e Projetos / In this master s thesis we approach the entanglement between quantum states as a resource to be used in the transmission of quantum information, particularly in quantum teleportation. Our focus was directed toward the study of partially entangled quantum states, that is, states in which the entanglement degree is below the maximum value allowed by quantum mechanics, since they are more likely to be found in actual devices than maximally entangled states. We review the properties and concepts that characterize vectors and operators from the viewpoint of linear algebra to formally present the postulates of quantum mechanics. We introduced the notion of qubits and showed how these can be applied to computing tasks via quantum logic gates.We explained in details the main quantum teleportation protocols that originated the methods used in this work, highlighting the advantages and disadvantages of each one. We presented protocols with new arrangements and introduced techniques that have improved some of the known methods, providing for those alternatives relating to construction and architecture in the arrangement of the channels. We proved that the probability associated with a multiple teleportation of a single qubit through channels in series tends to the maximum limit for a special case. We showed that imperfect states can be corrected and that it is possible to transmit a single qubit or a pair of qubits using GHZ states and modification of Bell basis states. We discussed the relationship between the final success probability value and the bases and channels degree of entanglement for each method. We concluded the thesis with a general analysis of all protocols, highlighting the most efficient ones with arguments that include theoretical and practical viewpoints. / Nesta dissertação abordamos o emaranhamento entre estados quânticos como um recurso a ser utilizado na transmissão de informação quântica, em especial no teletransporte quântico. Nosso foco foi direcionado para o estudo de estados quânticos parcialmente emaranhados, isto é, estados nos quais o teor de emaranhamento está aquém do máximo valor permitido pela mecânica quântica, visto que são mais prováveis de serem encontrados em dispositivos reais do que seriam estados maximamente emaranhados. Revisamos as propriedades e conceitos que caracterizam vetores e operadores sob a ótica da álgebra linear para em seguida apresentarmos formalmente os postulados da mecânica quântica. Introduzimos a noção de qubits e mostramos como estes podem ser aplicados em tarefas computacionais através de portas lógicas quânticas. Explicamos com detalhes os principais protocolos de teleporte quântico que originaram os métodos utilizados neste trabalho, destacando as vantagens e desvantagens de cada um. Apresentamos protocolos com arranjos inéditos e introduzimos técnicas que aperfeiçoaram alguns dos métodos de teleporte conhecidos, fornecendo alternativas para estes em relação à construção e arquitetura na disposição dos canais. Provamos que a probabilidade associada ao teleporte múltiplo de um qubit através de canais em série tende ao limite máximo para um caso especial. Mostramos que estados imperfeitos podem ser corrigidos e que é possível transmitir um qubit isolado ou um par de qubits empregando estados GHZ e modificações de estados da base de Bell. Discutimos a relação entre o valor da probabilidade de sucesso final e o grau de emaranhamento das bases e dos canais para cada método. Encerramos a dissertação com uma análise geral sobre todos os protocolos, destacando os mais eficientes com argumentos que englobam pontos de vista tanto teórico como prático.

One of the main goals of the theory of open quantum systems is to devise methods which help preserve the quantum properties of a system interacting with its environment. One possible pathway to achieve this goal is to use non-Markovian reservoirs, characterized by information backflows and revivals of certain quantum properties. These reservoirs usually require advanced engineering techniques, which may turn their implementation impractical. In this dissertation we propose an alternative technique: the injection of a classical colored noise, which induces the desired quantum non-Markovianity. In order to do that, we investigate the dynamics of a quantum system interacting with its surrounding environment and under the injection of a classical stochastic colored noise. A time-local master equation for the system is derived by using the stochastic wave function formalism and functional calculus. Afterwards, the non-Markovianity of the evolution is detected by using the Andersson, Cresser, Hall and Li measure, which is based on the decay rates of the master equation in canonical Lindblad-like form. Finally, we evaluate the measure for three different colored noises and study the interplay between environment and noise pump necessary to induce quantum non-Markovianity, as well as the energy balance of the system. / Um dos objetivos principais da teoria de sistemas quânticos abertos é desenvolver métodos que ajudem a preservar as propriedades quânticas de um sistema interagindo com o ambiente. Um possível caminho para alcançar essa meta é usar reservatórios não-Markovianos, caracterizados por refluxos de informação e renascimento de certas propriedades quânticas. Esses reservatóris geralmente requerem o uso de técnicas avançadas de engenharia, o que pode tornar sua implementação impraticável. Nessa dissertação nós propomos uma técnica alternativa: a injeção de um ruído colorido clássico, o qual induz a desejada não-Markovianidade quântica. De modo a fazer isso, nós investigamos a dinâmica de um sistema quântico interagindo com o ambiente e sob a injeção de um ruído colorido clássico estocástico. Uma equação mestra local no tempo é derivada usando-se do formalismo da função de onda estocástica e de técnicas de cálculo funcional. Após isso, a não-Markovianidade da evolução é detectada através da medida de Andersson, Cresser, Hall e Li, a qual é baseada nos coeficientes da equação mestra na forma de Lindblad-like canônica. Finalmente, nós calculamos a medida para três diferentes ruídos coloridos e estudamos a relação entre o ambiente e o bombeio estocástico necessária para induzir não-Markovianidade quântica, assim como o balanço de energia do sistema.

O processamento de informação quântica requer medidas, muitas vezes precedidas devoluções unitárias. Uma descrição realista de um computador quântico também deve levar em conta que o sistema interage com um ambiente externo - distinto do observador - que o remove de sua evolução ideal, gerando erros. Neste trabalho, fazemos um estudo da dinâmica de sistemas quânticos observados múltiplas vezes ou continuamente, enquanto interagem com ambientes externos. Para tanto, empregamos uma equação mestra híbrida, que permite modelar uma interação contínua e markoviana do sistema com o medidor, enquanto o ruído do ambiente apresenta características não markovianas. O estudo da dinâmica de uma medida contínua ruidosa revela que o sistema melhor preserva suas populações iniciais quando é realizada a medida de uma observável que não comuta com os operadores do ruído produzido pelo ambiente. Estes resultados, já conhecidos para o caso simples de um qubit de memória interagindo com o vácuo, são generalizados para uma temperatura inicial superior a zero e para um qubit submetido a uma porta quântica. A universalidade destes fenômenos de preservação da população inicial permite fazer analogia com o efeito Zenão quântico. Mantendo o mesmo formalismo, mas adaptando a interação com o ambiente para descrever um decaimento verificamos que o efeito Zenão quântico é observado para acoplamentos fracos com o ambiente. Tratamos também de como tal conhecimento sobre a preservação das populações pela medida auxilia na elaboração de melhores formas de preservar a informação em códigos quânticos. Com o auxílio da teoria das medidas fracas, propomos um possível método experimental simples para o teste da validade dos modelos de descrição de medidas contínuas. Com este estudo da dinâmica de uma medida quântica, esperamos elucidar questões de ordem prática no processamento de informação quântica, assim como ajudar no melhor entendimento de questões fundamentais, como o postulado da medida. / The processing of quantum information requires measurements, often preceded by unitary evolutions. A faithful description of a quantum computer should also take into account that the system interacts with an external environment - other than the observer - that removes it from its ideal evolution, causing errors. Here, we study the dynamics of quantum systems observed multiple times or continuously, while they interact with external environments. To do this, we employ a hybrid master equation, which allows us to model a continuous, Markovian interaction between the system and the measurement apparatus, while the environmental noise presents non-Markovian features. This study of the dynamics of the noisy continuous measurement reveals that the system better preserves its initial populations when the observable measured does not commute with the environmental noise operators. These results, already known for the simpler case of a memory qubit interacting with vacuum, are generalized for an initial temperature above zero and a qubit undergoing a quantum gate. The universality of these phenomena of preservation of the initial populations allows an analogy with the Quantum Zeno Effect. Keeping the same formalism, but adapting the environmental interaction to describe a decay, we verify that the quantum Zeno effect is observed for weak coupling with the environment. We also deal with how the knowledge about the preservation of the populations by the measurement helps in creating better ways to preserve the information in quantum codes. With the help of the weak measurement theory, we propose a simple experimental method to test the validity of models that describe a continuous measurement. With this study of the dynamics of a quantum measurement, we hope to help solve practical issues in quantum information processing, as well as provide greater insight into fundamental questions, such as the measurement postulate.

[pt] A criptografia quântica foi proposta como uma solução para o problema da distribuição de chaves criptográficas com segurança total garantida pelos princípios da mecânica quântica. Através dessa técnica é possível saber se um espião tentou interceptar a transmissão, o que é impossível utilizando técnicas de transmissão clássicas. Nesse trabalho foi feito um breve resumo da teoria de criptografia quântica, de suas técnicas de transmissão e dos problemas tecnológicos enfrentados. Foi analisada em detalhes a técnica de transmissão de qubits utilizando codificação de freqüência e feita uma comparação dos diferentes esquemas de modulação frente aos protocolos BB84 e B92. Foi demonstrado que os dois esquemas de modulação existentes (AM-AM e PM-PM) são na realidade equivalentes e foi proposto um novo esquema, o AM-PM o único que suporta o protocolo BB84 clássico. Medidas foram realizadas classicamente nos formatos AM-AM e AM-PM. / [en] Quantum cryptography has been proposed as a solution to the cryptographic key distribution problem with absolute security guaranteed by the principles of quantum mechanics. Through this scheme it is possible to find out whether a spy tried to eavesdrop on the transmission, which was impossible to discover using classical transmission techniques. In this work a brief review of quantum cryptography theory, transmission techniques and technological problems involved were performed. It was analyzed in detail the transmission technique employing frequency coding, and a comparison was made between the different modulation schemes and the BB84 and B92 protocols. It was demonstrated that the two existing modulation formats (AM-AM and PM-PM) are in fact equivalent and a new format (AM-PM) was proposed, the only one able to accommodate classical BB84. Classical measurements were performed on the AM-AM and AMPM formats.

Je souhaite réaliser un modèle théorique optimal pour les protocoles de partage de secret quantique basé sur l'utilisation des états graphes. Le paramètre représentatif d'un partage de secret à seuil est, entre autres la taille du plus grand ensemble de joueurs qui ne peut pas accéder au secret. Je souhaite donc trouver un famille de protocoles pour laquelle ce paramètre est le plus petit possible. J'étudie également les liens entre les protocoles de partage de secret quantique et des familles de courbes en géométrie algébrique. / I want to realize an optimal theoretical model for quantum secret sharing protocols based on graph states. The main parameter of a threshold quantum secret sharing scheme is the size of the largest set of players that can not access the secret. Thus, my goal is to find a collection of protocols for which the value of this parameter is the smallest possible. I also study the links between quantum secret sharing protocols and families of curves in algebraic geometry.

O processamento de informação quântica requer medidas, muitas vezes precedidas devoluções unitárias. Uma descrição realista de um computador quântico também deve levar em conta que o sistema interage com um ambiente externo - distinto do observador - que o remove de sua evolução ideal, gerando erros. Neste trabalho, fazemos um estudo da dinâmica de sistemas quânticos observados múltiplas vezes ou continuamente, enquanto interagem com ambientes externos. Para tanto, empregamos uma equação mestra híbrida, que permite modelar uma interação contínua e markoviana do sistema com o medidor, enquanto o ruído do ambiente apresenta características não markovianas. O estudo da dinâmica de uma medida contínua ruidosa revela que o sistema melhor preserva suas populações iniciais quando é realizada a medida de uma observável que não comuta com os operadores do ruído produzido pelo ambiente. Estes resultados, já conhecidos para o caso simples de um qubit de memória interagindo com o vácuo, são generalizados para uma temperatura inicial superior a zero e para um qubit submetido a uma porta quântica. A universalidade destes fenômenos de preservação da população inicial permite fazer analogia com o efeito Zenão quântico. Mantendo o mesmo formalismo, mas adaptando a interação com o ambiente para descrever um decaimento verificamos que o efeito Zenão quântico é observado para acoplamentos fracos com o ambiente. Tratamos também de como tal conhecimento sobre a preservação das populações pela medida auxilia na elaboração de melhores formas de preservar a informação em códigos quânticos. Com o auxílio da teoria das medidas fracas, propomos um possível método experimental simples para o teste da validade dos modelos de descrição de medidas contínuas. Com este estudo da dinâmica de uma medida quântica, esperamos elucidar questões de ordem prática no processamento de informação quântica, assim como ajudar no melhor entendimento de questões fundamentais, como o postulado da medida. / The processing of quantum information requires measurements, often preceded by unitary evolutions. A faithful description of a quantum computer should also take into account that the system interacts with an external environment - other than the observer - that removes it from its ideal evolution, causing errors. Here, we study the dynamics of quantum systems observed multiple times or continuously, while they interact with external environments. To do this, we employ a hybrid master equation, which allows us to model a continuous, Markovian interaction between the system and the measurement apparatus, while the environmental noise presents non-Markovian features. This study of the dynamics of the noisy continuous measurement reveals that the system better preserves its initial populations when the observable measured does not commute with the environmental noise operators. These results, already known for the simpler case of a memory qubit interacting with vacuum, are generalized for an initial temperature above zero and a qubit undergoing a quantum gate. The universality of these phenomena of preservation of the initial populations allows an analogy with the Quantum Zeno Effect. Keeping the same formalism, but adapting the environmental interaction to describe a decay, we verify that the quantum Zeno effect is observed for weak coupling with the environment. We also deal with how the knowledge about the preservation of the populations by the measurement helps in creating better ways to preserve the information in quantum codes. With the help of the weak measurement theory, we propose a simple experimental method to test the validity of models that describe a continuous measurement. With this study of the dynamics of a quantum measurement, we hope to help solve practical issues in quantum information processing, as well as provide greater insight into fundamental questions, such as the measurement postulate.

<p>Optimization and Machine learning (ML) have emerged as two positively disruptive methodologies and have thus resulted in unprecedented applications in several domains of technology. In recent years, ML has forayed into physical sciences and provided promising outcomes thanks to its ability in representing and generalizing complex functions to reveal underlying relations among variables describing a system. By casting ML as an optimization task, we first focus on its application in solving quantum many-body problems. Leveraging the power of quantum computation, we develop hybrid quantum machine learning protocols and implement benchmark tests to calculate the band structures of two-dimensional materials. We also show how this method can be used to estimate the critical point for a quantum phase transition. One  hurdle in such techniques is related to parameter optimization, wherein to obtain the desired result, the parameters have to be optimized, which can be computationally intensive. For a particular class of problem and a choice of algorithm, we deduce a simple parameter setting rule. This rule is projected as a heuristic and is validated numerically for several problem instances. Finally, by venturing into thermal photonics, a framework that takes advantage of the spectral and spatial information of hyperspectral thermal images to establish a completely passive machine perception, titled HADAR is presented. A conventional deep neural network is developed that utilizes the governing equation of HADAR and its performance in semantic segmentation is demonstrated. Altogether, this report establishes the need for creative algorithms that exploit modern hardware to solve complex problems that were previously deemed unsolvable.</p>

<p>La distribution quantique de clés est une technique cryptographique permettant l'échange de clés secrètes dont la confidentialité est garantie par les lois de la mécanique quantique. Le comportement particulier des particules élémentaires est exploité. En effet, en mécanique quantique, toute mesure sur l'état d'une particule modifie irrémédiablement cet état. En jouant sur cette propriété, deux parties, souvent appelées Alice et Bob, peuvent encoder une clé secrète dans des porteurs quantiques tels que des photons uniques. Toute tentative d'espionnage demande à l'espion, Eve, une mesure de l'état du photon qui transmet un bit de clé et donc se traduit par une perturbation de l'état. Alice et Bob peuvent alors se rendre compte de la présence d'Eve par un nombre inhabituel d'erreurs de transmission.</p> <p>L'information échangée par la distribution quantique n'est pas directement utilisable mais doit être d'abord traitée. Les erreurs de transmissions, qu'elles soient dues à un espion ou simplement à du bruit dans le canal de communication, doivent être corrigées grâce à une technique appelée réconciliation. Ensuite, la connaissance partielle d'un espion qui n'aurait perturbé qu'une partie des porteurs doit être supprimée de la clé finale grâce à une technique dite d'amplification de confidentialité.</p> <p>Cette thèse s'inscrit dans le contexte de la distribution quantique de clé où les porteurs sont des états continus de la lumière. En particulier, une partie importante de ce travail est consacrée au traitement de l'information continue échangée par un protocole particulier de distribution quantique de clés, où les porteurs sont des états cohérents de la lumière. La nature continue de cette information implique des aménagements particuliers des techniques de réconciliation, qui ont surtout été développées pour traiter l'information binaire. Nous proposons une technique dite de réconciliation en tranches qui permet de traiter efficacement l'information continue. L'ensemble des techniques développées a été utilisé en collaboration avec l'Institut d'Optique à Orsay, France, pour produire la première expérience de distribution quantique de clés au moyen d'états cohérents de la lumière modulés continuement.</p> <p>D'autres aspects importants sont également traités dans cette thèse, tels que la mise en perspective de la distribution quantique de clés dans un contexte cryptographique, la spécification d'un protocole complet, la création de nouvelles techniques d'amplification de confidentialité plus rapides à mettre en œuvre ou l'étude théorique et pratique d'algorithmes alternatifs de réconciliation.</p> <p>Enfin, nous étudions la sécurité du protocole à états cohérents en établissant son équivalence à un protocole de purification d'intrication. Sans entrer dans les détails, cette équivalence, formelle, permet de valider la robustesse du protocole contre tout type d'espionnage, même le plus compliqué possible, permis par les lois de la mécanique quantique. En particulier, nous généralisons l'algorithme de réconciliation en tranches pour le transformer en un protocole de purification et nous établissons ainsi un protocole de distribution quantique sûr contre toute stratégie d'espionnage.</p> <p>Quantum key distribution is a cryptographic technique, which allows to exchange secret keys whose confidentiality is guaranteed by the laws of quantum mechanics. The strange behavior of elementary particles is exploited. In quantum mechnics, any measurement of the state of a particle irreversibly modifies this state. By taking advantage of this property, two parties, often called Alice and bob, can encode a secret key into quatum information carriers such as single photons. Any attempt at eavesdropping requires the spy, Eve, to measure the state of the photon and thus to perturb this state. Alice and Bob can then be aware of Eve's presence by a unusually high number of transmission errors.</p> <p>The information exchanged by quantum key distribution is not directly usable but must first be processed. Transmission errors, whether they are caused by an eavesdropper or simply by noise in the transmission channel, must be corrected with a technique called reconciliation. Then, the partial knowledge of an eavesdropper, who would perturb only a fraction of the carriers, must be wiped out from the final key thanks to a technique called privacy amplification.</p> <p>The context of this thesis is the quantum key distribution with continuous states of light as carriers. An important part of this work deals with the processing of continuous information exchanged by a particular protocol, where the carriers are coherent states of light. The continuous nature of information in this case implies peculiar changes to the reconciliation techniques, which have mostly been developed to process binary information. We propose a technique called sliced error correction, which allows to efficiently process continuous information. The set of the developed techniques was used in collaboration with the Institut d'Optique, Orsay, France, to set up the first experiment of quantum key distribution with continuously-modulated coherent states of light.</p> <p>Other important aspects are also treated in this thesis, such as placing quantum key distribution in the context of a cryptosystem, the specification of a complete protocol, the creation of new techniques for faster privacy amplification or the theoretical and practical study of alternate reconciliation algorithms.</p> <p>Finally, we study the security of the coherent state protocol by analyzing its equivalence with an entanglement purification protocol. Without going into the details, this formal equivalence allows to validate the robustness of the protocol against any kind of eavesdropping, even the most intricate one allowed by the laws of quantum mechanics. In particular, we generalize the sliced error correction algorithm so as to transform it into a purification protocol and we thus establish a quantum key distribution protocol secure against any eavesdropping strategy.</p>