Multiple antennas help in combating the destructive effects of fading as well as improve the spectral efficiency of a communication system. Receive diversity techniques like maximal ratio receive combining have been popular means of introducing multiple antennas into communication systems. Space-time block codes present a way of introducing transmit diversity into the communication system with similar complexity and performance as maximal ratio receive combining. In this thesis we study the performance of space-time block codes in Rayleigh fading channel. In particular, the quasi-static assumption on the fading channel is removed to study how the space-time block coded system behaves in fast fading. In this context, the complexity versus performance trade-off for a space-time block coded receiver is studied. As a means to improve the performance of space-time block coded systems concatenation by convolutional codes is introduced. The improvement in the diversity order by the introduction of convolutional codes into the space-time block coded system is discussed. A general analytic expression for the error performance of a space-time block coded system is derived. This expression is utilized to obtain general expressions for the error performance of convolutionally concatenated space-time block coded systems utilizing both hard and soft decision decoding. Simulation results are presented and are compared with the analytical results. / Master of Science

Um material composto, formado pela mistura de muitos elementos, pode ser convenientemente descrito, a partir da forma com que a radiação interage com ele, como se fosse formado por um elemento ctício com um número atômico efetivo, Zef. Esse parâmetro não é constante com a energia, entretanto, pode ser considerado uma ferramenta útil na caracterização de tecidos biológicos, materiais tecido-equivalentes e dosímetros. Várias formas de determinação do Zef foram propostas na literatura, dentre elas, estão os métodos de atenuação que se baseiam no uso da seção de choque total (obtido a partir do coeciente de atenuação linear mássico (µ/)) e os métodos de espalhamento que se baseiam na razão entre as seções de choque diferenciais Rayleigh e Compton. Neste trabalho, buscou-se estudar métodos de obter Zef de forma experimental e teórica, por dois métodos (atenuação e espalhamento) para completar lacunas existentes na literatura. No método de atenuação, foi utilizado o (µ/) como grandeza de interação que, no conhecimento dos autores, ainda não foi utilizado. Para isso, arranjos experimentais que permitam determinar a densidade, , e o coeciente de atenuação linear, µ, foi construído. O µ (de materiais tecido-equivalentes e tecidos biológicos) foi determinado tanto para energia de 59.54 keV (fonte 241Am) com detector CdTe, quanto para faixa de energia entre 15 e 45 keV (tubo de raios X com alvo de W) com detector SDD. Um novo programa computacional para determinar o número atômico efetivo a partir do µ/ foi implementado e validado. A sensibilidade do método foi estudada de forma a determinar Zef de forma adequada. No método de espalhamento, arranjo experimental para detectar os fótons espalhados Rayleigh e Compton (utilizando fonte de 241Am e detector CdTe) foi construído, otimizado e validado. Um programa computacional para determinação de Zef através da razão Rayleigh/Compton, R/C, foi elaborado e validado . A sensibilidade do método foi estudada e analisada de forma a determinar Zef de maneira adequada. No método de atenuação, o arranjos para determinação do µ/ possibilitaram determiná-lo com diferenças menores que 6% quando comparados com a literatura e incertezas de 3.8% para 59.54 keV e até 7% na faixa de 15 à 45 keV. Também foi vericado que o método é adequado para determinar Zef para energias de até 60 keV, pois a partir deste valor, as incertezas em Zef aumentam (para mais de 10%). No método de espalhamento, o arranjo para determinação da razão Rayleigh/Compton possibilitaram obter medidas de R/C com menos de 10% de diferença com a literatura e incertezas de 7% e foi vericado ainda, que a faixa de momento transferido entre 1 Å1 e 2 Å1 é adequada para determinar Zef (incertezas menores que 3% de incerteza) / A composite material, formed by the mixture of several elements, can be conveniently described by, from the way the radiation interacts with it, as if it were formed by a factitious element with an eective atomic number (Zef). This parameter is not constant with the energy, however, can be considered as an useful tool for characterization of biological tissues, tissue-equivalent materials and dosimeters. Several ways for the determination of Zef were proposed by the literature, among them, are the attenuation methods which are based on the total cross section (derived form the mass attenuation coecient (µ/)) and the scatter methods which are based on the ratio between the Rayleigh and Compton dierential cross sections. In this work, were study ways to obtain Zef experimentally and theoretically by two methods (attenuation and scattering) to ll the gaps in the literature. In the attenuation method µ/ was used as interaction coecient which, in the authors knowledge, it hasnt been used. For this purpose, experimental arrangements to determine the density () and the linear attenuation coecient (µ) were built. The µ (of tissue-equivalent material and biological tissues) were determined both for 59.54 keV (241Am source) with a CdTe detector and for the energy rage between 15 and 45 keV (X ray tube, W target) with a SDD detector. A new computational program for determining Zef through µ/ was implemented and validated. The method sensibility was studied in the way to determine Zef properly. In the scattering method, an experimental arrangement to detect the Rayleigh and Compton scattered photons (with 241Am source and a CdTe detector) was built, optimized and validated. A computational program to determine Zef through Rayleigh to Compton ratio, R/C, was elaborated and validated. The methods sensibility was studied and analyzed to determine Zef properly. In the attenuation method, the arrangements for determining µ/ allowed to determine it with dierences smaller than 6% when compared with the literature and uncertainties of 3.8% for 59.54 keV and up to 7% in the range of 15 to 45 keV. It has also been found that the method is suitable for determining Zef for energies of up to 60 keV, because above that energy, the uncertainties in Zef increases (more than 10%). In the scattering method, the arrangement for determining R/C made it possible to obtain measurements of R/C with less than 10% dierence with the literature and uncertainties of 7% and it was veried that the momentum transfer range between 1 Å1 and 2 Å1 is suitable for determining Zef (uncertainties less than 3 % uncertainty)

A utilização de laser para estudo, análise e caracterização de materiais é de imensa importância no desenvolvimento e progresso da ciência e tecnologia, melhorando ciclicamente a área de ótica não linear (ONL). Estudos do comportamento de compostos orgânicos cujas estruturas são π conjugadas mostram sua ampla aplicabilidade em dispositivos eletrônicos e na área medicinal. A adição de grupos laterias doadores ou aceitadores de elétrons têm a capacidade de aumentar os efeitos ópticos não lineares nestes compostos, otimizando suas propriedades. Nesta dissertação de mestrado, foram realizados estudos em compostos orgânicos focados em suas características ópticas não lineares. Nove derivados de chalconas e quatro derivados de azometinas foram analisadas através da técnica de espalhamento hiper-Rayleigh (HRS) e da técnica de varredura-z, cujos efeitos óticos não lineares observados correspondem à segunda e terceira ordem da polarizabilidade, respectivamente. Com o espalhamento hiper-Rayleigh, foram determinadas a primeira hiperpolarizabilidade βHRS cujo valor mais alto foi 24×10−30 cm5/esu para as chalconas e 273×10−30 cm5/esu para as azometinas. A técnica de varredura-z foi utilizada para determinar o espectro da seção de choque de absorção de dois fótons (σ2PA) cujos valores de pico das moléculas derivadas de chalconas são de aproximadamente 15 GM e algumas apresentaram estados excitados que são permitidos apenas por absorção de dois fótons. Duas das moléculas derivadas de azometinas apresentaram valores de pico da ordem de 12 GM, uma não apresentou absorção de dois fótons e uma apresentou valor de pico de 55 GM. Além disso, foi utilizado o método de Soma-Sobre-Estados (SOS - do inglês Sum-Over-States) para determinar os momentos de dipolo de transição e diferença de momentos de dipolo permanente de todos os derivados. / The use of lasers for study, analyses and characterization of materials are of great importance for progress and development of science and technology, improving in a cyclic way the nonlinear optics area (NLO). Studies of the behave of π-conjugated organic compounds show its wide applicability in electronic devices and in the medicinal area. Adding electron donor or acceptor lateral groups to these compounds may alter its nonlinear optical effects, optimizing its properties. In this Master dissertation, studies on organic compounds focused in its nonlinear optical characteristics were realized. Nine chalcone derivatives and four azomethines derivatives were analyzed using the Hyper-Rayleigh Scattering technique (HRS) and Z-Scan technique, by which the nonlinear optical effects observed corresponds to the second and third order of the polarizabilities, respectively. With the HRS the first hyperpolarizability βHRS was determined, which the highest values was of 24×10−30 cm5/esu for the chalcones, and 273×10−30 cm5/esu for the azomethines. The Z-Scan technique was utilized for determining the two-photon cross-section σ2PA which peak values of the chalcone derivatives were of approximately 15 GM and some of them presented excited states that are only allowed by two-photon absorption. Two of the azomethines derivatives molecules presented peak values of the order of 12 GM, one of them did not present two-photon absorption, and one presented peak value of 55 GM. Furthermore, the Sum-Over-States (SOS) method was used to determine the transition dipole momentums and the difference of permanent dipole momentum of all of the derivatives.

Made available in DSpace on 2016-04-04T18:31:21Z (GMT). No. of bitstreams: 1 Fernando Lino.pdf: 1954006 bytes, checksum: a697088d7a7477d92d2183541ef1e1c6 (MD5) Previous issue date: 2008-02-08 / The literature describes the fast flat fading propagation model in mobile environment as a Rayleigh distribution with different standard deviations which considers only one parameter to describe the channel fading. In the Weibull distribution there are two parameters to be modeled: shape factor and scale factor. The Rayleigh distribution is a particular case of the Weibull distribution where the shape factor has a firm value equals to two. This work has the objective to evaluate, through practical measurements, both parameters in the Weibull distribution at an indoor environment. The results will show the real value of the shape factor parameter and the scale factor as well, indicating that the Weibull distribution is more flexible to describe a fast flat fading channel at an indoor propagation. / A literatura descreve a propaga??o fast flat fading em sistemas m?veis como o modelo de distribui??o Rayleigh com diferentes desvios padr?es, o qual considera apenas um par?metro para descrever o canal de desvanecimento. Na distribui??o de Weibull existem dois par?metros a serem modelados: fator de forma e fator de escala. A distribui??o Rayleigh ? um caso particular da distribui??o de Weibull onde o fator de forma tem um valor igual a dois. Este trabalho tem o objetivo de avaliar, atrav?s de medidas pr?ticas, ambos os par?metros da distribui??o de Weibull em um ambiente indoor. Os resultados ir?o mostrar os reais valores dos par?metros de forma e de escala, indicando que a distribui??o de Weibull ? mais flex?vel para descrever um canal flat fast fading em uma propaga??o indoor.

Este trabalho teve como objetivo a realização de um estudo sobre o método MASW (Análise Multicanal de Ondas de Superfície) avaliando-se comparativamente resultados obtidos por diferentes fontes (ativas e passivas), ondas Rayleigh e Love e diferentes modos de propagação das ondas de superfície. Dois ensaios sísmicos foram executados: com geometria de aquisição bidimensional e geofones triaxiais de 10Hz, e com geometria linear e geofones de componente vertical de 4,5Hz. Foram realizados estudos com as fontes marreta, compactador de solo, ruído ambiental e com o tráfego de veículos, a última através da técnica Passive Roadside MASW. Resultados de inversões com dados da componente vertical (ondas Rayleigh) das ondas de superfície foram avaliados com os de inversões conjuntas com a componente radial (ondas Rayleigh) e transversal (ondas Love). Analisou-se também os produtos de inversões da curva de dispersão do modo fundamental com os de inversões conjuntas com o primeiro modo superior. Os estudos foram realizados em frente ao Instituto de Astronomia, Geofísica e Ciência Atmosféricas (IAG) localizado no interior do campus Cidade Universitária Armando de Salles Oliveira (CUASO) da Universidade de São Paulo (USP) no bairro do Butantã, São Paulo. A área de estudo possui informações de sondagem mista com descrição geológica do material e valores SPT (Standard Penetration Test) que foram utilizados para validação dos resultados. O compactador de solo demonstrou ser uma melhor fonte ativa em relação a marreta através de diferentes aspectos: geração de maior energia tanto na componente vertical quanto na transversal; espectros (V,f) de melhor qualidade; os dados apresentam a vantagem de poderem ser processados através da técnica f-k beamforming. A aquisição com o ruído ambiental não possibilitou a interpretação de curvas de dispersão devido ao pouco tempo de aquisição utilizado. Os dados obtidos pela técnica Passive Roadside MASW contribuíram com os dados de ativa através do registro de frequências mais baixas. Além disso, devido a clara identificação do 1° modo superior em seu espectro (V,f) foi possível a identificação do mesmo modo no espectro (V,f) dos dados de ativa, interpretação até então duvidosa. No entanto, a inversão conjunta das curvas de dispersão dos dados Passive Roadside com as dos dados de ativa não resultou em uma melhor inversão comparada com a inversão obtida apenas pelas curvas de ativa. A inversão conjunta de curvas das componentes radial (ondas Rayleigh) e transversal (ondas Love) com as curvas obtidas da componente vertical (ondas Rayleigh) também não trouxe um melhor resultado quando comparada com a inversão alcançada apenas com as curvas da componente vertical. A utilização do primeiro modo superior com o modo fundamental, no entanto, mostrou trazer melhoras significativas nos resultados das inversões em comparação com inversões apenas da curva do modo fundamental. Considerando as incertezas envolvidas os melhores resultados deste trabalho são convergentes com os dados de sondagem da área de estudo. No atual estágio de desenvolvimento do método MASW diversas etapas são bastante dependentes do operador. Neste sentido os estudos realizados neste trabalho contribuem para um melhor entendimento do método nos seus fundamentos, parâmetros de aquisição e processamento. / This work aim to study the MASW (Multichannel Analysis of Surface Waves) method by comparing results obtained with different sources (active and passive), Rayleigh and Love waves and different modes of surface waves propagation. Two seismic tests were performed: one with two-dimensional acquisition geometry and 10Hz triaxial geophones, and the other with linear geometry and 4.5 Hz vertical component geophones. Studies were carried out with the following sources: sledgehammer, rammer compactor, ambient noise and vehicular traffic, the last through the Passive Roadside MASW technique. Inversions results with vertical component data (Rayleigh waves) were evaluated through joint inversions with the radial (Rayleigh waves) and transversal (Love waves) components. It were also analyzed the inversions results of the fundamental mode of the dispersion curve with the results of joint inversions with the first higher mode. The studies were carried out in front of the Instituto de Astronomia, Geofísica e Ciência Atmosféricas (IAG) (Institute of Astronomy, Geophysics and Atmospheric Science) located inside the university campus Cidade Universitária Armando de Salles Oliveira (CUASO) of Universidade de São Paulo (USP) (University of São Paulo) in the neighborhood of Butantã, São Paulo. The study area has information with a geological material description and SPT (Standard Penetration Test) values that were used to validate the results. The rammer compactor showed to be a better active source in relation to sledgehammer through different aspects: generation of greater energy in vertical and transverse components; better quality of (V,f) spectrum; the data have the advantage that they can be processed using the f-k beamforming technique. The acquisition with ambient noise did not allow the dispersion curves interpretation due to the short acquisition time used. Passive Roadside MASW data contributed to the active data through the lower frequency. Besides that, due to the clear identification of the 1st higher mode in its (V,f) spectrum it was possible to identify the same mode in the (V,f) spectrum of the active data, interpretation that was, until then, doubtful. However, the joint inversion of the Passive Roadside dispersion curves with the active curves did not produce better results compared to the inversion obtained only by the active curves. The joint inversion of dispersion curves from radial (Rayleigh) and transversal (Love waves) components with the curves obtained by the vertical component (Rayleigh waves) also did not bring a better result when compared with the inversion achieved only by the vertical component curves. The use of the first higher mode with the fundamental mode, however, showed significant improvements in the joint inversions results compared to inversions only of the fundamental mode curve. Considering the uncertainties involved, the best results of this work converge with the a priori information of the study area. At the current MASW method stage of development, several steps are quite dependent on the operator. Therefore, the studies carried out in this work contribute to a better understanding of the method in its fundamentals, acquisition parameters and processing.

L’emploi de matériaux composites à matrice organique (CMO) pour la réalisation de structures aéronautiques « tièdes », peut exposer ces matériaux à l’action d’environnements agressifs, qui peuvent entrainer des phénomènes de vieillissement et de dégradation sévères associés à la diffusion d’espèces au sein du réseau macromoléculaire des matrices polymères. La complexité de la microstructure des CMO utilisés pour ces applications peut complexifier la compréhension de phénomènes de dégradation. Le vieillissement humide des CMO préoccupe particulièrement les industriels du secteur aéronautique ; la diffusion de l’eau dans la matrice polymère du composite peut entrainer des phénomènes de gonflement hygroscopique, des modifications des propriétés mécaniques. Des méthodes expérimentales existent pour la caractérisation de ces phénomènes et pour l’identification des paramètres associés : ces méthodes préconisent souvent l’emploi d’éprouvettes saturées en humidité,ce qui nécessite de longs temps de conditionnement et un nombre élevé d’échantillons. Cette thèse a pour but d’établir des protocoles d’identification rapide des propriétés diffuso-mécaniques de matériaux polymères et CMO pour applications aéronautiques. La démarche mise en place dans cette thèse s’organise autour de quatre chapitres. Le premier chapitre présente une étude bibliographique sur les outils de modélisation des couplages diffuso-mécaniques et sur les méthodes de caractérisation des propriétés diffuso-mécaniques. L’étude bibliographique permet de préciser le cadre de travail, qui prévoit l’emploi d’un modèle diffuso-mécanique faiblement couplé où la diffusion d’eau suit la loi de Fick et le comportement mécanique est hygroélastique linéaire, dépendant de la concentration en eau. Le second chapitre présente la mise en place et le développement d’une méthode d’identification rapide des propriétés de diffusion anisotrope, adaptée à des CMO à architecture complexe.La méthode s’appuie sur des mesures de prise de masse d’échantillons de CMO, l’anisotropie de diffusion de ces matériaux est obtenue par rotation des axes principaux d’orthotropie. La méthode proposée représente une extension de la « méthode de la pente » introduite par Shen et Springer pour l’identification des propriétés de diffusion de matériaux orthotropes, et basée sur l’exploitation des courbes gravimétriques aux temps courts. A travers cette méthode, les coefficients principaux et les axes principaux d’orthotropie peuvent être identifiés. Une discussion sur les conditions d’équivalence de la diffusion 3D à la diffusion 1D en fonction de l’épaisseur de l’échantillon est également présentée à la fin de ce chapitre. Le troisième chapitre explore à travers une étude numérique la possibilité d’identifier rapidement les propriétés mécaniques affectées par la concentration en eau de matériaux polymères à travers des essais mécaniques sur plaques minces avec des gradients de concentration en eau. Des essais de traction et de flexion sont considérés. Pour des matériaux isotropes, dans un cadre hygroélastique, il est montré que cette méthode permet l’identification du module d’Young et du coefficient de Poisson dépendant de la concentration en eau avec un gain remarquable de temps d’essai par rapport à des essais sur des échantillons saturés en humidité. Enfin, le dernier chapitre propose à travers une étude numérique une méthode d’identification rapide des propriétés diffuso-mécaniques de matériaux isotropes basée sur l’emploi de plaques sollicitées par un champ asymétrique de concentration en eau. L’identification est ainsi effectuée à partir du suivi des déflexions engendrées par les champs de concentration. Le coefficient de dilatation hygroscopique et le module d’Young dépendant de la concentration en eau peuvent être identifiés durant l’essai, à l’état transitoire du conditionnement, avec un gain remarquable du temps d’essai par rapport à des essais sur échantillons saturés en humidité. / The use of organic matrix composite materials (OMC) for the realization of "warm" aeronautical structures, may expose these materials to aggressive environments: wet or gaseous environments,high temperatures, which may promote severe aging and degradation phenomena related to species diffusion within the macromolecular network of the polymer matrices. The complexity of the OMC microstructure used for these applications can complicate the understanding of degradation phenomena : for example, species diffusion can be isotropic, orthotropic or anisotropic, depending on the texture of the fibrous reinforcement. Humid aging of OMC is of particular concern for the aeronautical industry ; the diffusion of water in the polymer matrix of the composite may promote hygroscopic swelling, changes in mechanical properties (stiffness, strength). Experimental methods exist for the characterization of these phenomena and for the identification of the associated parameters : these methods often recommend the use of moisture saturated specimens, which require long conditioning times, sometimes often a relevant number of samples (1 sample for each saturated state), high costs. The aim of this work is to establish protocols for fast identification of the diffusomechanical properties of polymers and polymer based OMC materials for aeronautical applications. The approach implemented in this thesis is organized in four chapters. The first chapter presents a bibliographic study on coupled diffuso-mechanics modeling tools and on methods of characterization/identification of diffuso-mechanical properties, more particularly for OMC for aeronautical applications. The bibliographic study allows specifying the framework of the present research, which foresees the employment of a weakly coupled diffuso-mechanical model, where water diffusion follows the Fick’s law and the mechanical behavior is linear hygroelastic, depending on water content. The second chap-ter presents the setting up and the development of a method for fast identification of anisotropic diffusion properties, suitable for OMC with complex architecture, such as for instance, 2D or 3Dwoven OMC. The method relies on mass-gain measures of OMC samples, the diffusion anisotropy ofthese materials is obtained by rotating the axes of orthotropy. The proposed method represents an extension of the "slope method" introduced by Shen and Springer in the 1970s for the identification of the diffusion properties of orthotropic materials (such as laminated composites), and is based onthe exploitation of gravimetric curves at short times. Through this method, the principal coefficients and the principal axes of orthotropy can be identified. A discussion about the transition from 3Dto 1D diffusion as a function of the sample geometry is also presented at the end of this chapter. The third chapter explores through a numerical study the possibility of identifying in a fast way the mechanical properties affected by moisture of polymeric materials by the use of mechanical tests on thin plates with water concentration gradients. Traction and bending tests are taken into account.For isotropic materials, in a hygroelastic setting, it is showed that this method allows identifyng the water concentration dependent Young’s modulus and the Poisson’s ratio with a remarkable time gain compared to tests on moisture saturated samples. Finally, the last chapter proposes through a numerical study a method for fast identification of the diffuso-mechanical properties of isotropic materials based on the use of plates loaded by an asymmetric water concentration field. The identification is thus carried out from the monitoring deflections generated by the concentration fields. The moisturedependent hygroscopic expansion coefficient and Young’s modulus can be identified during the test,by exploiting the transient state of conditioning, with a remarkable time gain compared with moisture saturated samples.

On étudie la propagation d’ondes acoustiques dans un milieu poreux obéissant à la théorie de Biotet contenant une distribution aléatoire de cavités cylindriques et celle d’une onde acoustique dans un fluide contenant une distribution polydisperse aléatoire de sphères poreuses. Dans le premier cas on utilise la généralisation de Conoir-Norris de la formule de Linton-Martin. En effet, elle permet de prendre en compte le phénomène de la conversion entre les trois ondes (deux longitudinales et une transversale) se propageant naturellement dans un milieu poreux de Biot contenant une distribution aléatoire de cavités. Des expressions analytiques sont trouvées pour les nombres d’onde effectifs des ondes cohérentes dans la limite de Rayleigh (régime basse fréquence). Les approximations des masses volumiques et modules des milieux hétérogènes sont fournies jusqu’à l’ordre de c2 en concentration. Le cas limite des cavités fluides aléatoires dans une matrice élastique est également discuté. Dans le deuxième cas on détermine les nombres d’ondes, modules et masses volumiques effectifs pour des distributions polydisperses de sphères poroélastiques. Pour y parvenir, les formules récentes du nombre d’onde effectif données par Linton et Martin dans le cas dilué monodisperse ont été modifiées. Compte tenu de l’incertitude entourant la prédiction de la distribution en taille des obstacles, trois densités de probabilité différentes sont étudiées et comparées : uniforme, Schulz et log-normal. Plus précisément, la limite de Rayleigh (régime de basse fréquence) est prise en compte lorsque les longueurs d’onde peuvent être supposées très grandes par rapport à la taille des obstacles. Dans cette limite, des formules simplifiées des concentrations sont fournies en fonction du paramètre caractérisant la dispersion en taille. / We study the propagation of acoustic waves in a porous medium obeying the Biot theory and containing a random distribution of cylindrical cavities and that of an acoustic wave in a fluid containing a random polydisperse distribution of porous spheres. In the first case we use the generalization by Conoir-Norris of the Linton-Martin formula. It allows to take into account the phenomenon of the conversion between the three waves (two longitudinal and one transverse) naturally propagating in a porous Biot medium containing a random distribution of cavities. Analytical expressions are found for the effective wavenumbers of coherent waves in the Rayleigh limit (low frequency regime). The approximations of the densities and heterogeneous media modules are presented up to the order of c2 in concentration. The limiting case of random fluid cavities in an elastic matrix is also discussed. In the second case the effective wavenumbers, moduli and density are determined for polydisperse distributions of poroelastic spheres. To achieve this, the recent formulas of the effective wave number given by Linton and Martin in the dilute monodisperse case have been modified. Given the uncertainty for predicting the distribution in size of obstacles, three different probability densities are studied and compared: uniform, Schulz and lognormal. More precisely, the Rayleigh limit is taken into account when the wave lengths can be assumed to be very large compared to the size of the obstacles. Within this limit, simplified formulas for concentrations are provided depending on the parameter characterizing the size dispersion.

Dans cette thèse, le compromis entre la consommation d'énergie et le délai dans les réseaux de capteurs sans fil est étudié en exploitant efficacement les liens non fiables. Nous proposons une métrique de l'efficacité énergétique et une métrique de délai moyenne qui sont combinés avec le modèle de lien radio à erreur. En utilisant ces deux métriques et un modèle réaliste de lien radio, nous déduisons la limite inférieure et le front de Pareto du compromis Energie-Délai sont calculés pour trois types de systèmes de communication: la communication multi-saut traditionnelle, la communication opportuniste et la communication coopérative MIMO (CMIMO) dans trois différents canaux: le canal a bruit additif blanc Gaussien, le canal a évanouissements rapides de Rayleigh et le canal a évanouissement de Rayleigh par blocs. Les expressions de ces limites faibles sont obtenues et vérifiées par des simulations bidimensionnelles de réseaux de capteurs distribués aléatoirement. Sur la base de l'expression de la borne inférieure du compromis Energie-Délai, un cadre multi-couche est fourni pour optimiser les paramètres des couches physique, MAC et routage sous contrainte de délai. Les résultats obtenus montrent que pour parvenir à une meilleure performance Energie-Délai, les mécanismes de communication suivants devraient être adoptés selon le type de canal : les communications multi-saut traditionnelles sont les plus performantes pour le canal à bruit additif blanc Gaussien, alors que les communications opportunistes le sont pour les canaux à évanouissements de Rayleigh par blocs et enfin le CMIMO pour les canaux à évanouissements rapides de Rayleigh.

In many systems, instabilities can lead to time-dependent behavior, and instabilities can act as mechanisms for sustained chaos; an understanding of the dynamical modes governing instability is thus essential for prediction and/or control in such systems. In this thesis work, we have developed an approach toward characterizing instabilities quantitatively, from experiments on the prototypical Rayleigh-Bénard convection system. We developed an experimental technique for preparing a given convection pattern using rapid optical actuation of pressurized SF6, a greenhouse gas. Real-time analysis of convection patterns was developed as part of the implementation of closed-loop control of straight roll patterns. Feedback control of the patterns via actuation was used to guide patterns to various system instabilities. Controlled, spatially localized perturbations were applied to the prepared states, which were observed to excite the dominant system modes. We extracted the spatial structure and growth rates of these modes from analysis of the pattern evolutions. The lifetimes of excitations were also measured, near a particular instability; a critical wavenumber was found from the observed dynamical slowing near the bifurcation. We will also describe preliminary results of using a state estimation algorithm (LETKF) on experimentally prepared non-periodic patterns in a cylindrical convection cell.

Le problème de l'initiation de l'explosion de vapeur est abordé ici en étudiant la fragmentation thermique (faibles perturbations en pression) d'une goutte chaude entourée d'un film de vapeur stable. La fragmentation semble consécutive à des contacts locaux entre la goutte et le réfrigérant. Néanmoins, le mécanisme exact qui déforme la goutte suite à ces contacts est incertain. Après une étude des ordres de grandeur mis en jeu, nous nous apercevons qu'un contact peut fragmenter la goutte en un temps très bref. Nous adoptons alors une approche consistant à utiliser comme critère d'explosion la simple réalisation d'un contact. Afin de valider cette approche, nous cherchons à retrouver les variations expérimentales des seuils d'explosion en fonction de la pression ambiante et en fonction de la température du réfrigérant. Le modèle retrouve les variations expérimentales, les seuils sont retrouvés avec une certaine incertitude. Le contact est obtenu par deux mécanismes rapprochant les liquides : une compression globale du film de vapeur due à la perturbation et une amplification des défauts de l'interface entre le réfrigérant et la vapeur. Il est apparu que c'est le mécanisme de compression globale qui explique principalement les variations expérimentales. Suite à ces résultats, nous avons effectué des extrapolations du modèle afin de nous approcher des conditions dans lesquelles une explosion de vapeur peut se produire industriellement (c'est à dire dans les réacteurs nucléaires à eau pressurisée).