Nous nous intéressons dans cette thèse à la relation originale et problématique qui s’est établie, dans la période qui va de l’après-guerre jusqu’à la fin des années 1970, entre, d’une part, la temporalité de l’écriture telle que vécue par l’écrivain et, d’autre part, la temporalité produite par le texte. Le schème paradoxal qui préside à cette relation pourrait se résumer ainsi : écrire pour savoir pourquoi écrire. Si la formule semble consonner avec la vulgate d’une écriture autotélique, elle nous semble plus profondément poser la question complexe du désir du scripteur, quand toute relation univoque entre le dedans et le dehors est mise en cause. Pour mener à bien cette étude, nous procédons en deux grands moments. Le premier aborde cette configuration par un bord théorique : nous suivons le fil du syntagme de « point aveugle », en ses acceptions métaphoriques et conceptuelles, dans un corpus essayistique extrait du grand moment théorique français. La seconde propose deux parcours de lecture dans les œuvres littéraires narratives fictionnelles de Bernard Pingaud et d’Henri Thomas – écrivains qui, selon des modalités distinctes, montrent deux voies d’articulation originale des registres temporels considérés. Aussi différents soient-ils, ils partagent ce principe : écrire pour savoir ce que, sans cette traversée, on ne pourrait savoir, en s’affrontant à cet irréductible point aveugle qui se donne comme condition de possibilité et d’impossibilité du sens / The focus of this thesis is the relationship - in the period between the end of World War Two and the late 1970s - between the temporality of writing as experienced by the writer, and the temporality produced by the text itself. The paradoxical structure of this relationship can be captured by the following formula: writing in order to know why one writes. Though this formula may seem to invoke the modernist received wisdom concerning autotelitic writing, this thesis takes it as the starting point in order to understand the desires of the “scriptor” (to use Roland Barthes’s term) when all unambiguous relations between the internal and the external are called into doubt. This study will proceed in two major parts. The first will apply a theoretical lens to the notion of the “blind spot” as it appears in its various metaphorical and conceptual senses in the essayistic corpus of so-called “French theory”. The second will consist of a literary analysis of the fictional narratives of Bernard Pingaud and Henri Thomas: writers who, each in their own distinct modalities, reveal an original way of articulating the forms of temporality at the center of this study. As different as they are, Pingaud and Thomas each share the aim of writing in order to understand that which, without writing, one could never understand. In pursuit of this aim, both writers come face-to-face with a blind spot that poses itself as the condition of both the possibility and the impossibility of meaning

Although women have slowly been accepted in America's top orchestras, they are still a minority. Certain instrument sections remain predominantly male in spite of the implementation of blind auditions. Inconsistency in the use of the screen in blind auditions is partially responsible for the lack of women in orchestras, but the problem is seeded in cultural constructions that act as barriers for female brass, percussion and string bass players. Gender constructions have dictated that women should not play instruments that are loud, heavy or require physical exertion. Nonetheless, many women have shown that they are capable of performing as well as men on these instruments. Brass players are often the highest paid in an orchestra, thus women musicians face economic discrimination when they face bias at the hiring level or in the workplace.

L'Internet des objets (ldO) connaît un intérêt grandissant depuis plusieurs années, avec la multiplication des objets dits « connectés » et des applications associées. Pour assurer la connectivité de ces objets, des réseaux sans fil longue portée à faible consommation énergétique, dédiés à l'ldO, ont été développés, ces réseaux proposant une connectivité adaptée aux contraintes propres de l'ldO. Plus récemment, le 3GPP a fait évoluer ses technologies 2G et 4G pour pouvoir également fournir des solutions de connectivité adaptées à l'ldO. Pour obtenir l'extension de couverture requise, ces évolutions utilisent entre autres un mécanisme de répétition en aveugle. Dans ce contexte, cette thèse s'intéresse aux mécanismes de recombinaison au sein du récepteur et à l'impact de l'évolution temporelle du canal de propagation et des imperfections du système sur les performances de ces mécanismes, en prenant la technologie 2G comme cas d'application. Tout d'abord, nous nous intéressons aux performances des différents mécanismes dans le cadre d'un récepteur sans imperfection. Puis, nous considérons une synchronisation fréquentielle imparfaite et une estimation de canal imparfaite. Dans ce dernier cas, de nouveaux schémas de recombinaison mixte, alliant deux mécanismes de recombinaison successifs, sont proposés. Enfin, la dernière partie de l'étude consiste en une implémentation du système complet sur des cartes radio-logicielles. Les résultats de cette thèse peuvent être exploités afin d'adapter le schéma de combinaison en fonction de l'application visée, et ainsi de réduire la consommation énergétique du système. / The Internet of Things (loT) is a concept, where a large number of connected devices are communicating together through the same network, constituting an autonomous and intelligent entity. The increasing number of connected devices and IoT services confirms the growing interest for the loT and its applications. To provide connectivity to the devices, several dedicated wireless low power and wide area networks have been created. Recently, with the Release 13, the 3GPP decided to modify the 2G and 4G technologies, to be able to also provide such connectivity for the loT devices, creating the field of Cellular-loT. These evolutions imply a coverage extension compared to the legacy technologies. To obtain this extension, these new standards especially use a blind repetition mechanism. In this context, this thesis studies the performance of several diversity combiners at the receiver, and observes the impact of the temporal evolution of the propagation channel and of imperfections in the receiver. The 2G GSM system is chosen as the application case. Firstly, we consider a receiver without imperfection. Secondly, we consider imperfect frequency synchronization in the receiver. Then, we consider imperfect channel estimation and propose a hybrid combining scheme, obtained by mixing two different combining mechanisms. Finally, in the last part of our study, we carry out a hardware implementation of the system into a software-radio platform. With the theoretical and simulation results provided in this thesis, one can better evaluate the potential of each combining scheme proposed in the framework of loT communications to efficiently benefit from blind repetition mechanisms.

Restoring vision to the blind has been an objective of several research teams for a number of years. It is known that spots of light -- phosphenes -- can be elicited by way of electrical stimulation of surviving retinal neurons. Beyond this, however, our understanding of prosthetic vision remains rudimentary. To advance the realisation of a clinically viable prosthesis for the blind, a versatile integrated circuit neurostimulator was designed, manufactured, and verified. The neurostimulator provides electrical stimuli to surviving neurons in the visual pathway, affording blind patients some form of patterned vision; besides other benefits (independence), this limited vision would let patients distinguish between day and night (resetting their circadian rhythm). This thesis presents the development of the neurostimulator, an interdisciplinary work bridging engineering and medicine. Features of the neurostimulator include: high-voltage CMOS transistors in key circuits, to prevent voltage compliance issues due to an unknown or changing combined tissue and electrode/tissue interface impedance; simultaneous stimulation using current sources and sinks, with return electrodes configured to provide maximum charge containment at each stimulation site; stimuli delivered to a two dimensional mosaic of hexagonally packed electrodes, multiplexing current sources and sinks to allow each electrode in the whole mosaic to become a stimulation site; electrode shorting to remove excess charge accumulated during each stimulation phase. Detailed electrical testing and characterisation verified that the neurostimulator performed as specified, and comparable to, or better than, other vision prostheses neurostimulators. In addition, results from several animal experiments verified that the neurostimulator can elicit electrically evoked visual responses. The features of the neurostimulator enable research into how simultaneous electrical stimulation affects the visual neural pathways; those research results could impact other neural prosthetics research and devices.

Thesis (Ph. D.)--University of Hong Kong, 2006. / Title proper from title frame. Also available in printed format.

Due to ubiquity of internet, the development of electronic commerce is growing up rapidly. Many payment mechanisms of electronic transactions are proposed, such as electronic cash, credit cards, and electronic wallets. Electronic cash (e-cash) is like the digital version of paper cash, which possesses some key features such as anonymity and untraceability, and it can be used on the Internet for various electronic transactions. In general, electronic cash is classified into two types which are on-line electronic cash and off-line electronic cash. Each of them is suitable for different applications and environments. All of the proposed electronic cash schemes only focus on one of two types. In these schemes, users must decide which type of electronic cash they will use later before withdrawing. However, in this thesis we propose a novel electronic cash scheme which supports a user to withdraw a generic e-cash and then decide to spend it as an on-line e-cash or an off-line e-cash. Our proposed scheme is more convenient for users and more flexible for the bank and shops. It not only satisfies the necessary properties of electronic cash but also provides flexibility.

This thesis attempts to render architectural agency and aesthetics within the typological discussion of the data center in the rural American landscape. The disciplinary question of the role of architecture and aesthetics in data center design is related to earlier examples of factories and warehouses during modernity. The data center alters the traditional representative role of architecture; they are massive, horizontal buildings that are only conceivable from an aerial perspective, driven by logistics and efficiency. This thesis engages these issues by focusing on the point at which the architectural and programmatic problems of the data center converge, the building form and envelope. This thesis engages the building envelope as an expanded surface that considers not only logistical and environmental issues, but also engages the social and political architectural questions related to the identity of the data center in the rural landscape.

In this thesis, a blind symbol timing synchronization algorithm based on cyclic prefix for OFDM systems in multipath fading channels is proposed. It finds the starting point of symbol timing for using appropriate weights which are designed from channel delay spread characteristics. In multipath fading channels, the conventional ML (maximum likelihood) algorithm estimate is biased and has a large variance due to the effect of channel delay spread. The proposed exponential weighting methods not only solve the above problems but also improve the accuracy of symbol timing. Particularly, the proposed method does not require the information of SNR and channel length. From computer simulation results, the proposed method outperforms the other conventional algorithms and is also robust against the effect of multipath fading channels.

In this thesis, we study the effect of channel estimation on the performance of distributed space-time coding (DSTC) in amplify-and-forward (AF) cooperative networks. The relay based transmission takes two phase. In phase I, the source transmits a block of symbols, which include training symbols and data to destination. After receiving signals at relay, the DSTC is adopted to re-encode signals in order to achieve diversity gain at relay nodes. At destination, the signals received in two phase are combined and used to detected data symbols. In the thesis, for AF cooperative networks, the signal received at destination is effected the multiplication of channel coefficients on the source to relay and relay to destination links. Before detection, channel coefficients of all links need to be estimated. We propose a semiblind method to estimate the channel coefficients of direct link and the relay links. The semi-blind channel estimation scheme, exploits a small number of training symbols and second-order statistics of received signals. To improve the detection quality, the channel estimation is modified by treating the detected symbols as extra training symbols. Through simulation, it shows that the proposed channel estimation and the modification leads to obvious performance improvement.

Due to resource-saving and efficiency consideration, electronic voting (e-voting) gradually replaces traditional paper-based voting in some developed countries. An anonymous e-voting system that can be used in elections with large electorates must meet various security requirements, such as anonymity, uncoercibility, tally correctness, unrecastability, verifiability, and so on. Especially, the uncoercibility property is an essential property which can greatly reduce the possibility of coercion and bribe. Since each voter can obtain one and only one voting receipt in an electronic voting system, coercers or bribers can enforce legal voters to show their voting receipts to identify whether the enforced voters follow their will or not. It turns out that the coercion and bribe will succeed more easily in digital environments than that in traditional paper-based voting. In this dissertation, we analyze four possible scenarios leading to coercion and discover that the randomization property is necessary to blind-signature-based e-voting systems against coercion. Based on this result, we extend our research and come up with two provably secure randomized blind signature schemes from different cryptographic primitives, which can be adopted as key techniques for an electronic voting system against coercion and bribery.