The doctoral thesis deals with spectrum sensing in cognitive radio networks (CRN). A number of international organizations are currently actively engaged in standardization of CRN and it points out to the fact that this technology will be widely used in the near future. One of the key features of this technology is a dynamic access to the spectrum, which can be affected by many different harmful factors occurring in the communication chain. The thesis investigates the influence of selected factors on the spectrum sensing process. Another contribution of the work is the optimization of the Kolmogorov - Smirnov statistical test that can be applied for the primary user signal detection. The work also incorporates the analysis of the influence of the harmful effects caused by the commonly used transmitters and receivers on various spectrum sensing methods. The investigations are verified by the results of the simulations and also by the measurements with experimental platforms based on the software-defined radio (SDR).

Analyzing the structure of a dataset is a challenging problem in high-dimensions as the volume of the space increases at an exponential rate and typically, data becomes sparse in this high-dimensional space. This poses a significant challenge to machine learning methods which rely on exploiting structures underlying data to make meaningful inferences. This dissertation proposes the <i>n</i>-TARP method as a building block for high-dimensional data analysis, in both supervised and unsupervised scenarios.<div><br></div><div>The basic element, <i>n</i>-TARP, consists of a random projection framework to transform high-dimensional data to one-dimensional data in a manner that yields point separations in the projected space. The point separation can be tuned to reflect classes in supervised scenarios and clusters in unsupervised scenarios. The <i>n</i>-TARP method finds linear separations in high-dimensional data. This basic unit can be used repeatedly to find a variety of structures. It can be arranged in a hierarchical structure like a tree, which increases the model complexity, flexibility and discriminating power. Feature space extensions combined with <i>n</i>-TARP can also be used to investigate non-linear separations in high-dimensional data.<br></div><div><br></div><div>The application of <i>n</i>-TARP to both supervised and unsupervised problems is investigated in this dissertation. In the supervised scenario, a sequence of <i>n</i>-TARP based classifiers with increasing complexity is considered. The point separations are measured by classification metrics like accuracy, Gini impurity or entropy. The performance of these classifiers on image classification tasks is studied. This study provides an interesting insight into the working of classification methods. The sequence of <i>n</i>-TARP classifiers yields benchmark curves that put in context the accuracy and complexity of other classification methods for a given dataset. The benchmark curves are parameterized by classification error and computational cost to define a benchmarking plane. This framework splits this plane into regions of "positive-gain" and "negative-gain" which provide context for the performance and effectiveness of other classification methods. The asymptotes of benchmark curves are shown to be optimal (i.e. at Bayes Error) in some cases (Theorem 2.5.2).<br></div><div><br></div><div>In the unsupervised scenario, the <i>n</i>-TARP method highlights the existence of many different clustering structures in a dataset. However, not all structures present are statistically meaningful. This issue is amplified when the dataset is small, as random events may yield sample sets that exhibit separations that are not present in the distribution of the data. Thus, statistical validation is an important step in data analysis, especially in high-dimensions. However, in order to statistically validate results, often an exponentially increasing number of data samples are required as the dimensions increase. The proposed <i>n</i>-TARP method circumvents this challenge by evaluating statistical significance in the one-dimensional space of data projections. The <i>n</i>-TARP framework also results in several different statistically valid instances of point separation into clusters, as opposed to a unique "best" separation, which leads to a distribution of clusters induced by the random projection process.<br></div><div><br></div><div>The distributions of clusters resulting from <i>n</i>-TARP are studied. This dissertation focuses on small sample high-dimensional problems. A large number of distinct clusters are found, which are statistically validated. The distribution of clusters is studied as the dimensionality of the problem evolves through the extension of the feature space using monomial terms of increasing degree in the original features, which corresponds to investigating non-linear point separations in the projection space.<br></div><div><br></div><div>A statistical framework is introduced to detect patterns of dependence between the clusters formed with the features (predictors) and a chosen outcome (response) in the data that is not used by the clustering method. This framework is designed to detect the existence of a relationship between the predictors and response. This framework can also serve as an alternative cluster validation tool.<br></div><div><br></div><div>The concepts and methods developed in this dissertation are applied to a real world data analysis problem in Engineering Education. Specifically, engineering students' Habits of Mind are analyzed. The data at hand is qualitative, in the form of text, equations and figures. To use the <i>n</i>-TARP based analysis method, the source data must be transformed into quantitative data (vectors). This is done by modeling it as a random process based on the theoretical framework defined by a rubric. Since the number of students is small, this problem falls into the small sample high-dimensions scenario. The <i>n</i>-TARP clustering method is used to find groups within this data in a statistically valid manner. The resulting clusters are analyzed in the context of education to determine what is represented by the identified clusters. The dependence of student performance indicators like the course grade on the clusters formed with <i>n</i>-TARP are studied in the pattern dependence framework, and the observed effect is statistically validated. The data obtained suggests the presence of a large variety of different patterns of Habits of Mind among students, many of which are associated with significant grade differences. In particular, the course grade is found to be dependent on at least two Habits of Mind: "computation and estimation" and "values and attitudes."<br></div>

The doctoral thesis deals with spectrum sensing and subsequent use of the frequency spectrum by multicarrier communication system, which parameters are set on the basis of the optimization technique. Adaptation settings can be made with respect to several requirements as well as state and occupancy of individual communication channels. The system, which is characterized above is often referred as cognitive radio. Equipments operating on cognitive radio principles will be widely used in the near future, because of frequency spectrum limitation. One of the main contributions of the work is the novel usage of the Kolmogorov – Smirnov statistical test as an alternative detection of primary user signal presence. The new fitness function for Particle Swarm Optimization (PSO) has been introduced and the Error Vector Magnitude (EVM) parameter has been used in the adaptive greedy algorithm and PSO optimization. The dissertation thesis also incorporates information about the reliability of the frequency spectrum sensing in the modified greedy algorithm. The proposed methods are verified by the simulations and the frequency domain energy detection is implemented on the development board with FPGA.

L’avènement des aimants permanents et les progrès récents dans l’électronique de puissance ont joué un rôle majeur dans l’évolution de la motorisation électrique des véhicules. Actuellement, les machines synchrones à aimants permanents (MSAP) grâce à leurs performances, et surtout leur efficacité énergétique, sont considérées comme les candidats idéaux pour les chaînes de traction des véhicules hybrides et électriques. Toutefois, en raison du vieillissement des matériaux, des défauts de fabrication ou des conditions de fonctionnement assez sévères, différents types de défauts sont capables de survenir dans les composants de la machine, ses organes de commande ou de mesure. Pour répondre aux exigences de sûreté, de fiabilité et de disponibilité, l’intégration d’une approche de surveillance et de diagnostic de défauts, dans le groupe motopropulseur électrique automobile, devient de plus en plus primordiale. Dans ce contexte, l’objectif de la thèse est de contribuer au diagnostic et à la caractérisation de défauts dans la MSAP par une analyse vibratoire. En premier temps, des approches analytiques de modélisation de la MSAP et des défauts : de court-circuit inter-spires, d’excentricité et de démagnétisation rotoriques serontproposées. L’intérêt majeur de tels modèles, dans le cadre du diagnostic, est d’étudier le comportement de la machine en présence de défauts étudiés afin d’en déduire les méthodes de détection les plus adaptées. En outre, des modèles numériques seront développés afin de les confronter aux parties magnétique et mécanique analytiques de la machine ainsi qu’au défaut de démagnétisation. Dans la phase d’analyse des impacts de défauts, nous allons nous focaliser sur les cas d’excentricité et de démagnétisation rotoriques. Les indicateurs de défauts seront extraits des représentations du signal vibratoire dans le temps et l’espace et de leurs transformées de Fourier, pour les cas de défauts simples et les cas de deux défauts combinés. Pour les cas simples, deux approches de localisation seront proposées : la première utilise le principe de tests statistiques et de tables de signatures, inspirée des méthodes de diagnostic à base de modèles, alors que la deuxième repose sur un banc de trois réseaux de neurones, où chacun est à une entrée et une sortie et destiné à localiser un type de défaut. Enfin, les performances des deux approches, en termes de robustesse et d’adaptabilité, seront comparées pour les mêmes ensembles de seuillage/d’apprentissage et de test. / The advent of new magnetic materials and recent advances in power electronics have played a major role in the progress of hybrid electric vehicles. Nowadays, permanent magnet synchronous machines (PMSM) thanks to their performances, especially their energy efficiency, are considered as ideal candidates for the traction chains of hybrid and electric vehicles. However, due to material aging, manufacturing defects or severe operating conditions, different types of faults are capable to occur in the machine components, its control or measuring devices. In order to ensure safety, reliability and availability, the integration of a fault diagnosis and condition monitoring approach in the automotive electrical powertrain system is becoming more and more important. In this context, the aim of the thesis is to contribute to the diagnosis and characterization of faults in the PMSM based on a vibration analysis. First, analytical modeling approaches for the PMSM and inter-turn short-circuits, eccentricity and rotor demagnetization faults will be proposed. The major interest of such models, in a diagnosis context, is to study the behavior of the machine in the presence of studied faults in order to deduce the most suitable detection methods. In addition, numerical models will be developed in order to validate the analytical magnetic and mechanical parts of the machine as well as the demagnetization fault. In the phase of fault impact analysis, we will focus on the cases of rotor eccentricity and demagnetization. The fault indicators will be extracted from the vibratory signal representations in time and space domains and their Fourier transforms, in the cases of single faults and the cases of two combined faults. For single fault cases, two diagnosis approaches will be proposed: the first uses the principle of statistical tests and fault signature tables, inspired by model-based diagnosis methods, while the second relies on a set of three neural networks, such as each one is with a single input and a single output and dedicated to isolate one type of fault. Finally, the performance of these two approaches, in terms of robustness and adaptability, will be compared for the same training and test sets.

Since the last few years, the unprecedented growth of communication systems involving the propagation of electromagnetic waves is particularly due to developments in mobile phone technology. The reverberation chamber is a reliable bench-test, enabling the study of the effects of electromagnetic waves on a specific electronic appliance. However, the operating of a reverberation chamber being rather complicated, development of numerical models are of utmost importance to determine the crucial parameters to be considered.This thesis consists in the modelling and the simulation of the operating principles of a reverberation chamber by means of the Finite-Difference Time-Domain method. After a brief study based on field and power measurements performed in a reverberation chamber, the second chapter deals with the different problems encountered during the modelling. The consideration of losses being a very important factor in the operating of the chamber, two methods of implementation of these losses are set out in this chapter. Chapter~3 consists in the analysis of the influence of the stirrer on the first eigenmodes of the chamber; the latter modes can undergo a frequency shift of several MHz. Chapter~4 shows a comparison of results issued from high frequency simulations and theoretical statistical results. The problem of an object placed in the chamber, resulting in a field disturbance is also tackled. Finally, in the fifth chapter, a comparison of statistical results for stirrers having different shapes is set out.