This thesis first presents a pole-zero placement algorithm for the systematic design of high-order phase-locked loops (PLL) serving as anti-imaging and anti-aliasing filters for time-mode signal processing applications. A 6th order PLL is designed and fabricated on a printed circuit board and is interfaced to a production mixed-signal tester. The correct filtering operation and large-signal transfer characteristic of the PLL are verified with an all-digital DFT solution. The digital test input is driven by a single clock, which can be programmed directly from an ATE high-speed digital pattern generator. As application of these high-order PLLs, an accurate and low-cost clock delay generation system is presented. With proper compensation and calibration, a delay resolution of 15 ps is achieved over an 8.4 ns range. This technique is shown experimentally to be a viable solution for clock alignment and for measuring jitter at a 50 GHz effective sampling rate. / Ce mémoire présente tout d'abord une approche systématique descendante pour la conception de boucles à verrouillage de phase (PLL) ayant un ordre arbitraire et opérant comme filtre anti-image ou anti-repliement pour le traitement de signal dans le domaine temporel. Un PLL de 6e ordre a été conçu et fabriqué sur une carte de circuit imprimé montée sur un tester à signaux-mixtes (ATE). La fonction de filtrage et la caractéristique de transfert de grands-signaux sont vérifiées à l'aide d'une solution de conception pour test (DFT) entièrement numérique. Le signal d'entrée numérique est cadencé par une horloge unique. Par conséquent, le signal de test peut être programmé sans effort à partir de l'instrument numérique à haute-vitesse (HSD) d'un testeur à signaux-mixtes (ATE). De plus, un système précis et économique de génération de délai d'horloge est présenté comme une application du PLL construit. A l'aide de calibration et compensation appropriées, une résolution de délai de l'ordre de 15 ps est réalisée pour un intervalle de 8.4 ns. Cette technique est démontrée expérimentalement comme étant une solution viable pour l'alignement d'horloge et pour mesurer le vacillement d'horloge à un rythme de sous-échantillonnage de 50 GHz.

Diffractive optical elements (DOEs) are an important component in the success of optical Microsystems. Electron beam lithography is a key part of fabricating these elements with submicron feature dimensions. This thesis presents work done on the development of a process for the fabrication of multilevel diffractive optics in glass substrates using this method. This project investigates various challenges involved in the process, addresses possible problems that may arise and proposes and investigates solutions to resolve them. Sources of possible error in the creation and transfer of the patterns are identified and methods of eliminating or minimizing these errors are presented. Some of the main sources of error arise from charging due to electron accumulation and alignment issues during electron beam lithography. / Éléments d'optiques diffractives (EODs) composent une partie essentielle dans le succès de microsystèmes optiques. Lithographie à faisceau d'électrons est un élément clé pour la fabrication des structures avec des dimensions critiques submicroniques. Cette thèse présente le travail fait sur le développement d'un processus pour la fabrication des optiques diffractives en utilisant cette méthode. Ce projet étudie des divers défis impliqués dans ce processus, traite des problèmes qui pourrait surgir et propose des solutions pour les résoudre. Les sources d'erreur possible dans la création et le transfert des modèles sont identifiées et des méthodes de les éliminer ou les minimiser sont présentées. Certaines des erreurs sont attribuées à l'accumulation d'électrons et aux problèmes d'alignement lors de la lithographie.

For realistic nanostructures, there are inevitably some degree ofdisorder such as impurity atoms, imperfect lattices, surfaceroughness, etc.. For situations where disorder locate randomly inthe nanostructure, any calculated quantum transport results shouldbe averaged over disorder distributions. A brute force approach isto generate many disorder configurations, calculate each of them,and then average the results. For atomistic first principlesmodeling, such a brute force averaging is computationallyprohibitive - if not impossible, to perform. It is therefore veryimportant and useful to develop a theoretical framework where thedisorder averaging is done analytically before atomic firstprinciples analysis is carried out.In this thesis, we have developed such a first principlesnon-equilibrium quantum transport theory and its associated modelingsoftware for predicting disorder scattering in nano-electronicdevices. Our theoretical formalism is based on carrying out densityfunctional theory (DFT) within the Keldysh non-equilibrium Green'sfunction (NEGF) framework, and a non-equilibrium vertex correction(NVC) theory for handling disorder configurational average at thenon-equilibrium density matrix level. In our theory, we use thecoherent potential approximation to calculate disorder averaging ofthe device Hamiltonian and one particle Green's functions, and useNVC to calculate correlated multiple impurity scattering at thenon-equilibrium density matrix level. After the NEGF-DFT-NVCself-consistent calculation is converged, we calculate thetransmission coefficients by a second, unavoidable, vertexcorrection. The NEGF-DFT-NVC theory allows us to predictnon-equilibrium quantum transport properties of nanoelectronicdevices with atomistic disorder from first principles without anyphenomenological parameters. The theory and implementation detailsare presented.We have applied the NEGF-DFT-NVC method to investigate severalimportant problems associated with disorder scattering innano-electronic device systems. These include interface roughnessscattering in Fe/vacuum/Fe magnetic tunnel junctions; the diffusivescattering of carriers due to oxygen vacancies in Fe/MgO/Fe magnetictunnel junctions; the surface roughness scattering that enhancesresistivity of copper interconnect wires; and effects of barrierlayer coating for Cu interconnects. Our investigations reveal veryimportant role played by the atomic level defects and impurities toboth equilibrium and nonequilibrium quantum transport properties,and results compare favorably with the corresponding experimentaldata. / Dans le cas de nanostructures concrètes, un certain degré de désordre apparaît inévitable tel que la présence d'impuretés, de structures cristallines imparfaites, de surfaces rugueuses, etc. Dans les situations où le désordre se matérialise aléatoirement dans la nanostructure, tout calcul de transport quantique devrait être réalisé en tant que moyenne sur plusieurs distributions désordonnées. Une approche par force brute consiste à générer plusieurs configurations désordonnées, calculer les propriétés d'intérêt pour chacune d'entre elles, et ensuite effectuer la moyenne des résultats. Dans le cas de la modélisation atomique à partir des principes premiers, une telle moyenne par force brute est prohibitive en terme de temps de calcul - sinon impossible. Il est ainsi très important et utile de développer un cadre théorique où la moyenne de désordre est faite analytiquement avant que l'analyse par les principes premiers ne soit effectuée. Dans cette thèse, nous avons développé une telle théorie de transport quantique hors équilibre à partir des principes premiers et le logiciel de modélisation associé pour la prédiction de la diffusion par désordre dans des dispositifs nanoélectroniques. Notre formalisme théorique est basé sur l'utilisation de la théorie de la fonctionnelle de densité (DFT) dans le cadre de la fonction de Green hors équilibre de Keldysh(NEGF), et sur l'emploi d'une correction de sommet hors équilibre (NVC) pour le traitement des moyennes configurationnelles de désordre au niveau de la matrice de densité hors équilibre. Dans notre théorie, nous utilisons l'approximation du potentiel cohérent afin de calculer les moyennes de désordre de l'Hamiltonien du dispositif et les fonctions de Green à une particule, et nous utilisons la NVC pour calculer la diffusion par impuretés multiples corrélée au niveau de la matrice de densité hors équilibre. Après que le calcul auto-cohérent NEGF-DFT-NVC ait convergé, nous calculons les coefficients de transmission par le biais d'une seconde correction de sommet inévitable. La théorie NEGF-DFT-NVC nous permet de prédire les propriétés de transport quantique hors équilibre de dispositifs nanoélectroniques avec désordre au niveau atomique à partir des principes premiers sans aucun paramètre phénoménologique. La théorie et les détails d'implémentation sont présentés dans ce travail. Nous avons appliqué la méthode NEGF-DFT-NVC afin d'examiner plusieurs problèmes importants associés à la diffusion par désordre dans des systèmes de dispositif nanoélectronique. Cela inclut la diffusion par rugosité de surface dans des jonctions tunnel magnétiques Fe/vide/Fe; la diffusion due à des lacunes d'oxygène dans des jonctions tunnel magnétiques Fe/MgO/Fe; la diffusion par rugosité de surface qui décuple la resistivité de fils deconnexion en cuivre; et les effets des revêtements couche barrière pour des connexionsen Cu. Notre étude révèle le rôle très important joué par les défauts de niveau atomique et les impuretés vis-à-vis des propriétés de transport quantique à la fois en équilibre et hors équilibre, et les résultats se comparent favorablement aux données expérimentales correspondantes.

Impacts of Geomagnetic storms on Trans-Canadian GridsInduced "dc" currents due to geomagnetic storms have caused power system blackouts and transformer damages in the past. Statistically, large geomagnetic storms follow a 10-11 year sunspot cycle, but geomagnetic disturbances with significant impacts can occur at any time. The geomagnetically induced currents (GIC) have wide range of detrimental effects on otherwise normal operation of power systems, communication systems, railway systems, and can cause oil and gas pipeline corrosion. This thesis presents a method of estimating the magnitudes of terrestrial voltages; which is based on solving the magnetic diffusion equation. The method predicts the oft-quoted figure of "1 volt per kilometer" for the range of geomagnetic field variations (magnitudes and frequencies) and ground resistivity frequently encountered. The prediction formula has been validated by the plane wave method. Geomagnetic storms have the tendency to disrupt the normal operation and even lead to major power outages depending upon the intensity of geomagnetic activity; whereas the major effects include transformer equipment damage, relay tripping, SVC failures resulting in minor or major power outages. / RÉSUMÉImpacts of Geomagnetic storms on Trans-Canadian GridsPar le passé, les courants induits produits par des orages électromagnétiques ont causés des dommages à des transformateurs ainsi que des pannes de courant majeures (blackouts). Statistiquement, les orages géomagnétiques sévères suivent un cycle d'activité solaire de 10 à 11 ans. Cependant, des perturbations géomagnétiques peuvent avoir un impact important et peuvent apparaître à n'importe quel moment. Les courants induits géomagnétiques ont un large champ d'effets néfastes sur les réseaux électriques, les systèmes de communication, les systèmes ferroviaires et peuvent causer de la corrosion dans les oléoducs. Cette thèse présente une méthode pour estimer l'amplitude des tensions terrestres qui est basée sur les équations de diffusion magnétique. Cette méthode prédit la figure couramment utilisée de « 1 volt par kilomètre » pour une étendue de champs géomagnétiques (amplitudes et fréquences) et de résistances du sol fréquemment rencontrées. La formule de prédiction a été validée par la méthode d'onde plane.Les orages géomagnétiques ont tendance à déranger les opérations normales et même à engendrer des pannes électriques majeures dépendamment de l'intensité de l'activité géomagnétique. Les principaux effets sont des dommages à l'équipement des transformateurs, des déclenchements de relais de protection, des défaillances des compensateurs statiques (SVC), ce qui causent parfois des pannes de courant mineures ou majeures.

This thesis tackles the problem of data conversion in the Residue Number System (RNS). The RNS has been considered as an interesting theoretical topic for researchers in recent years. Its importance stems from the absence of carry propagation between its arithmetic units. This facilitates the realization of high-speed, low-power arithmetic. This advantage is of paramount importance in embedded processors, especially those found in portable devices, for which power consumption is the most critical aspect of the design. However, the overhead introduced by the data conversion circuits discourages the use of RNS at the applications. In this thesis, we aim at developing efficient schemes for the conversion from the conventional representation to the RNS representation and vice versa. The conventional representation can be in the form of an analog continuous-time signal or a digital signal represented in binary format. We present some of the currently available algorithms and schemes of conversion when the signal is in binary representation. As a contribution to this field of research, we propose three different schemes for direct conversion when interaction with the real analog world is required. We first develop two efficient schemes for direct analog-to-residue conversion. Another efficient scheme for direct residue-to-analog conversion is also proposed. The performance and the efficiency of theses converters are demonstrated and analyzed. The proposed schemes are aimed to encourage the utilization of RNS in various real-time and practical applications in the future. / Cette thèse aborde le problème de la conversion de données dans le système numérique de résidus (Residue Number System - RNS). Le système RNS a été considéré comme un sujet intéressant par de nombreux chercheurs ces dernières années. Son importance découle de l'absence de la propagation de retenue entre ses unités de calcul. Ceci facilite la réalisation de circuits arithmétiques à grande vitesse et de faible puissance. Cet avantage est d'une importance primordiale dans les processeurs embarqués, en particulier ceux qu'on retrouve dans les appareils portables, pour lesquels la consummation d'énergie est l'aspect le plus critique de la conception. Cependant, le traitement supplémentaire introduit par les circuits de conversion de données décourage l'utilisation du RNS au niveau des applications. Dans cette thèse, nous cherchons des schémes efficaces pour la conversion de la représentation conventionnelle à la représentation RNS et vice-versa. La représentation conventionnelle peut être sous la forme d'un signal analogique en temps continu où d'un signal échantillonné numérique représenté en format binaire. Nous présentons quelques algorithmes actuellement disponibles et les systèmes de conversion associés lorsque le signal est sous unereprésentation binaire. Dans notre contribution à ce domaine de recherche, nous proposons trois astuces différentes pour la conversion lorsqu'une interaction avec le monde analogique réel est nécessaire. Nous dévelopons deux systèmes efficaces pour la conversion directe du domaine analogique à RNS. Un autre système efficace pour la conversion directe de RNS à analogique est également proposé. La performance et l'efficacité de ces convertisseurs sont mises en évidence et analysées. Les schémas proposés sont destinés à encourager l'utilisation du RNS dans diverses applications dans l'avenir.

This thesis aims to build the groundwork for a distributed network of collaborating, intelligent video surveillance cameras. This work is implemented on a system of low-cost embedded-microprocessor-based camera modules. Each camera develops a person detection classifier using the Winnow algorithm for unsupervised, online learning. Training examples are automatically extracted and labelled from the current frame, and the classifier is then applied to the image to detect and locate person instances. / To improve detection accuracy, multiple cameras with overlapping fields of view collaborate to confirm their results. We present a novel, unsupervised calibration technique that allows each camera module to understand its spatial relationship with the other cameras. During runtime operation, cameras efficiently apply the learned spatial correlations to confirm detections. This method implicitly handles non-overlapping field of view regions that cannot be confirmed by other cameras. Its computational efficiency makes it well-suited to real-time processing on our hardware.

In this thesis, we present two new methods to overcome the effects of both camera calibration errors and partial emptiness of surface voxels in voxel-coloring. Voxel-coloring is a relatively new volumetric method for 3D scene reconstruction from multiple calibrated views. The quality of reconstruction is affected by the presence of errors in the estimated calibration parameters. Furthermore, a voxel forming a scene surface may be partially empty as there is no prior knowledge about the scene surface. Both of these sources of error result in outlier pixels in voxel projections in the input images. These outlier pixels affect the photo-consistency test of the voxel and tend to result in over-carving of the reconstructed 3D scene. The existing methods to handle these errors are either insufficient or too complex. We propose a method to handle the effect of calibration errors and call it Adaptive Gaussian Averaging. It makes use of the information of the error probability distribution of projected pixel coordinates due to camera parameter errors. We propose another method to reduce the effects of partial emptiness of surface voxels and we call it Area Weighting. In this method we use the pixel count in voxel projections to weight the projections in voxel's color statistics calculations. Our proposed methods are simple and can be incorporated into the existing voxel-coloring algorithms easily. We also conduct experiments on our own calibrated data sets to verify the effectiveness of the proposed methods. The experimental results show that both of our proposed methods have the ability to improve the results of existing algorithms. We also compare the results of our proposed methods with the results of an existing method that handle these errors too, the r-Consistency. We find that our proposed methods have the ability to adapt to the level of errors present in the system, and perform better than r-Consistency when the effect of these errors is higher on voxel-coloring.

This thesis presents the design and experimental results of a 7.3GHz notch image reject filter, combined with a 5.8GHz low-noise amplifier (LNA), for integrated heterodyne receiver front-ends. A new image reject filter implementation is proposed. Q-enhancement circuitry for on-chip inductors are used to optimize the depth of image rejection. Experimental results show that more than 62dB of image rejection at 7.3GHz can be obtained in a standard CMOS 0.18mum technology, while operating from a 1.8V supply. The LNA exhibits a gain of 15.8dB and an IIP3 of -5.3dBm while consuming 9mW of power. With maximum image rejection, the LNA-notch combination circuit achieves a 4.1dB noise figure at 5.8GHz. The proposed notch filter alone can operate from a 1V supply voltage. It is shown analytically how circuit stability can be ensured. / The implementation of new robust and stable high-Q CMOS image reject filters, which enables the realization of fully integrated heterodyne 5GHz RF receivers is also presented. A cascade of two notch filters with their image reject frequencies slightly offsetted is proposed, in order to obtain a wide image rejection bandwidth, without having to resort to the overhead of automatic tuning circuitry. Thus, power consumption, area, and complexity are significantly reduced. Experimental results show that more than 30d$ of image rejection can be obtained in a standard 0.18mum CMOS technology, over a 400MHz bandwidth centered at 7.4GHz.

Future wireless communication systems will have to support large uplink/downlink data rates within the allocated frequency spectrum in order to support emerging multi-media applications. Recently explored multiple antenna or multi-input multi-output (MIMO) systems have the potential of achieving large increases in wireless channel capacities and spectral efficiencies. However, powerful channel codes and sophisticated signal processing techniques are required to achieve these increases for low signal-to-noise ratios and in the presence of multipath fading. Motivated by the third generation wire-less standard and efforts towards fourth generation systems, this dissertation explores turbo-coded uplink transmission over Rayleigh faded MIMO channels, when channel state information is available only at the receiver (base-station). In this context, unlike previously proposed MIMO signal detection techniques of exponential or polynomial complexity, we first derive a novel soft-decision MIMO signal detector, which is of linear complexity in the number of transmit antennas, receive antennas and users. / The proposed detector is integrated into a new iterative receiver for turbo-coded MIMO systems and simulation results are presented for various antenna configurations on slow, frequency non-selective Rayleigh fading channels. The explored MIMO systems achieve target bit error rate (BER) performance of 10-5 within 1.5 to 2.9 dB of the ergodic capacity limit in single and multi-user scenarios. Furthermore, system performance is shown to be robust to the effects of spatial correlation, channel estimation errors and imperfect power control. In addition, turbo-coded digital recording systems with multiple heads and tracks were considered as a further application of the proposed detector and iterative receiver. Good BER performance is achieved in these systems despite strong intertrack and intersymbol interference. / Finally, we extend the union bound based method from turbo codes on the Gaussian noise channel to the proposed turbo-coded systems on wireless MIMO channels. Calculated bounds on BER performance are within 0.5 to 0.8 dB of the simulated results in the error floor region. In addition, a new performance analysis technique is proposed, which provides BER performance bounds within 0.1 to 0.5 dB of the simulated performance in the waterfall region. Furthermore, the latter technique can also provide accurate bounds for selected coded digital recording systems.

New digital communication applications, such as multimedia, require very powerful error correcting codes that deliver low error rates while operating at low to moderate ratios (SNRs). Turbo codes have reasonable complexity and can achieve very low error rates if a proper interleaver design is in place. The use of well-designed interleavers result in very low error rates, especially for medium to long interleavers where turbo codes offer the greatest potential for achieving high minimum distance (d min) values. / The reliable determination of a code's error performance at very low error rates using simulations may take months or may not be practical at all. However, the knowledge of dmin and its multiplicities can be used to estimate the error rates at high SNR. This thesis is concerned with efficient and accurate distance measurement methods for turbo codes. Since high values of dmin can be caused by high input weight values, say up to 20 or higher, if a brute force algorithm is used the accurate determination of dmin requires that all possible input sequences of input weight up to 20 be tested. Testing all possible input sequences becomes impractical as the size of the interleaver and the value of input weight increase. Thus, the accurate determination of the distance spectrum, or at least dmin and its multiplicities, is a significant problem, especially for interleavers that yield high dmin. Based on Garello's true distance measurement method, this thesis presents an efficient and accurate distance measurement method for single- and double-binary turbo codes that uses proper trellis termination such as dual-termination or tail-biting. This method is applied to determine the distance spectra for the digital video broadcasting with return channel via satellite (DVB-RCS) standard double-binary turbo codes. It is also used to design new interleavers for DVB-RCS that yield a significant improvement in error performance compared to those in the standard. / This method fits particularly well with tail-biting turbo codes that use structured interleavers. This is because the distance properties repeat and this method can use this knowledge to reduce the search space. The reduction in search space results in significant reduction in complexity (i.e., execution time), which allows the determination of high dmin values in reasonable time. This efficiency is demonstrated for both single- and double-binary turbo codes, using structured interleavers that have high dmin values for various code rates. This method reduces the execution tunes by a factor of 40 to 400.