Dans cette thèse, nous proposons l'analyse numérique et le développement d'algorithmes partitionnés pour coupler l'écoulement du sang dans différents comparti- ments cardiovasculaires (3D-3D, 3D-0D) Dans une première partie, un problème couplé fluide-fluide est introduit. Sur l'interface qui sépare les domaines, des conditions aux limites de type Robin-Robin dérivées de la formulation d'interface de Nitsche sont considérées. Nous proposons différents schémas explicites dont la stabilité est analysée dans la norme de l'énergie. Des simulations numé- riques illustrent le potentiel des méthodes présentées. La deuxième partie propose des applications cardiovasculaires plus réalistes. Tout d'abord, un modèle d'ordre réduit pour les valves cardiaques est décrit. Sans traiter l'inter- action fluide-structure avec le sang, les valves sont remplacées par des surfaces agissant comme des résistances immergées dans le fluide. Des simulations numériques montrent l'efficacité et la robustesse de ce modèle. Pour finir, une formulation ALE est utilisée pour la résolution d'un modèle fluide sur un domaine mobile. Nous montrons qu'en ajoutant un terme consistent, une inégalité d'éner- gie stable peut être obtenue sans considérer aucune hypothèse de Loi de Conservation Géométrique. Le travail se termine avec des simulations numériques sur la dynamique du sang dans le ventricule gauche, couplé avec l'écoulement du sang dans l'aorte. / In this thesis we present the numerical analysis and the development of parti- tioned algorithms in order to couple the blood dynamics in different cardiovascular compart- ments (3D-3D, 3D-0D). In the first part a fluid-fluid coupled problem is introduced. On the interface between the domains Robin-Robin boundary conditions, derived from the interface Nitsche’s formulation, are considered. We suggest different staggered explicit schemes whose stability is analyzed in the energy norm. Extensive numerical experiments illustrate the accuracy of the methods presented. The second part deals with more realistic cardiovascular applications. First a reduced order model for the heart valves is described. Without dealing with fluid-structure interaction with the blood flow, the valves are replaced by immersed surfaces acting as resistances on the fluid. Numerical simulations show the efficiency and the robustness of this model in the framework of a fluid-fluid interaction scheme. In the end, an ALE formulation is used to solve a fluid model in a moving domain. We show that adding a suitable consistent term, a stable energy inequality can be obtained without considering any Geometric Conservation Laws. The work ends with numerical sim- ulations on blood dynamics in the left ventricle coupled with the blood flowing in the aorta.

Yes / The world is experiencing an explosion in demand for ultra-high data rates with far greater expectations in the next few years. These expectations, given the bandwidth-demanding applications such as augmented and virtual reality and other beyond-5G applications, motivate the exploration of higher-frequency communication in the terahertz (THz) bands. However, THz communication is faced with many technical challenges, primarily due to the high susceptibility to blockages that limit its applications. Here, reconfigurable intelligent surfaces (RIS) provide alternative paths to circumvent such blockage effects and ensure reliable, spectral, and energy-efficient communication, thus advancing the THz-RIS technology concept. However, the ambitious targets of ubiquitous and global connectivity can only be satisfied by many technologies extending to multiple domains, from terrestrial networks to non-terrestrial network (NTN) domains. The use of airborne and spaceborne networks is considered a potential solution for addressing these challenges due to their dynamism, coverage, and ability to leverage their altitude for achieving line-of-sight communication for enhanced signal quality and network performance. Therefore, unmanned aerial vehicles, high-altitude platform stations, and satellites are poised to use flying THz-based RISs to improve air-to-ground and space-to-ground communication reliability while exploiting novel RIS architectures, techniques and enablers to address the issues regarding the propagation conditions, hardware limitations, network complexity and system performance. The aim in this paper is to present the discussion and a survey on the technical advances on THz-RIS NTNs, in addition to outlining potential applications, architectural variants, influencing properties, as well as its prospects, associated challenges, open issues and future directions towards high-data rate THz-RIS NTN communication for 6G and beyond. / This work was supported in part by the Universiti Kebangsaan Malaysia through Dana Impak Perdana 2.0 under Grant DIP-2022-020; and in part by the Engineering and Physical Sciences Research Council [grant number EP/Z001544/1] through the UK Research and Innovation (UKRI)-funded Marie Skłodowska-Curie Actions (MSCA) Postdoctoral Fellowship

Navigation is integral to modern infrastructure, with GPS serving as the foundation for applications in transportation, banking, and communications. Despite its widespread success, GPS is vulnerable to failures due to its low received signal power, susceptibility to jamming, and reduced accuracy in dense urban environments and deep fades. A failure of GPS could have severe consequences, making it crucial to explore alternative or supplementary navigation technologies. This work investigates the potential of three approaches—low Earth orbit (LEO) satellites, reconfigurable intelligent surfaces (RISs), and near-field propagation—to enhance localization accuracy and resilience. LEO satellites, originally designed for communication, have recently seen widespread deployment through constellations such as Starlink, OneWeb, and Kuiper. Their growing presence presents an opportunity to explore their feasibility for 9D localization, which includes 3D position, velocity, and orientation estimation. However, using LEO satellites for localization introduces significant challenges, including ionospheric delays, high Doppler shifts, limited synchronization due to the absence of atomic clocks, and uncertainty in satellite ephemeris data. To address these challenges, we leverage estimation theory and the Fisher Information Matrix (FIM) to establish theoretical bounds on localization performance. Our analysis shows that localization is possible using signals from multiple LEO satellites observed across several time slots, even in the presence of time and frequency offsets. We derive closed-form expressions for the FIM and identify conditions under which localization is feasible, highlighting the required number of satellites, base stations, and transmission slots. RISs provide another avenue for enhancing localization by dynamically shaping wireless propagation channels through software-controlled meta-material surfaces. We analyze the localization potential of RISs under both near-field and far-field conditions, focusing on angle of incidence, reflection, and orientation estimation. Our FIM-based study reveals that in far-field scenarios, angle estimation is only feasible with multiple RIS phase profiles, whereas in near-field, a single phase profile suffices. This distinction has implications for RIS-aided positioning, indicating that a single RIS reflection may not provide sufficient information for localizing a user in the far-field unless additional mechanisms, such as multiple reflections or phase variations, are employed. We further investigate near-field propagation, assessing the available localization information when a source transmits to a destination node. Our Fisher information analysis reveals that in the near-field regime, 3D orientation and position can be jointly estimated, whereas in the far-field, only 2D orientation and position can be determined. Additionally, we explore the impact of propagation model mismatches on direction-of-arrival (DOA) estimation using the MUSIC algorithm. Our simulations quantify the performance degradation when incorrect assumptions are made about the propagation environment, showing that estimation accuracy suffers significantly when near-field effects are ignored. Notably, in near-field scenarios, using a far-field-based beamforming model leads to an underestimation of DOA estimation errors, while in the far-field, MUSIC remains effective with appropriate beamforming design. Overall, our findings indicate that LEO satellites, RISs, and near-field propagation hold significant potential for overcoming the limitations of GPS and enabling precise localization for next-generation applications. By leveraging these technologies, it may be possible to achieve robust navigation in environments where GPS performance is compromised, paving the way for resilient and high-accuracy positioning solutions. / Doctor of Philosophy / Navigation is so ubiquitous that we spend little time thinking about its inner workings. The Global positioning system (GPS) works so well that navigation for airplanes and cars relies heavily on it. Moreover, on the less-known side, the use of GPS in the banking sector is a little-known, but vitally important use-case as banking institutions utilize GPS to provide strict timing to synchronize all their clocks. Considering these, a GPS failure would be catastrophic, costing millions and possibly resulting in a humongous loss of life. Hence, it is essential to address the scenarios where GPS can fail. The main challenge that GPS faces is the low received power of its signal. It is well known that the received signal power of GPS signals is similar to viewing a light bulb from a distance of 1 km. This dramatically increases the possibility of the signal being jammed. Also, it can become unreliable in deep fades and urban canyons since the signal is received well below the noise floor. Due to these concerns, there is a need to provide a backup to replace or, at the very least, augment GPS. Three technologies that could enable navigation are reconfigurable intelligent surfaces (RISs), near-field signal processing, and low Earth orbit (LEO) satellites. Interest in LEO satellites has skyrocketed due to the launch of several new constellations - Boeing, SpaceMobile, OneWeb, Telesat, Kuiper, and Starlink as well as the deployment of satellites into existing constellations - Orbcomm, Iridium, and Globalstar. Although these LEO satellites are being deployed for communication purposes, their ubiquitous deployments mandate a rigorous investigation of the utility of these satellites to provide 9D localization (3D positioning, 3D velocity estimation, and 3D orientation estimation) of a terrestrial re- ceiver. We utilize estimation theory to show that with sufficient spacing between time slots and in the presence of time and frequency offsets and Doppler rate, it is possible to perform 9D localization (3D position, 3D velocity, and 3D orientation estimation) of a receiver by utilizing the signals from three LEO satellites observed during three transmission time slots received through multiple receive antennas. RISs are envisioned as planar surfaces comprising sub-wavelength-sized meta-materials ca- pable of controlling a wireless propagation channel by applying a desired transformation on the incoming signal through the software control of each meta-material, and this unique ability to control the harsh wireless channel has led to many works investigating the suit- ability of RISs for localization. We show under certain conditions that we can estimate the angles of incidence, reflection, and the orientation offset at the RIS and utilize these angles for localization. We also investigate the available information during near-field propagation under the case of a downlink source transmitting to a destination node and show that the near-field provides additional information for localization. Our work indicates that the RISs, LEO satellites, and near-field propagation could be used to achieve the accuracy needed for futuristic localization use cases.

Un des avantages des lasers femtosecondes de haute puissance (TW-PW) est de pouvoir obtenir, au foyer d'une optique focalisante, des intensités très élevées atteignant jusqu'à 10^22W.cm^-2 soit un champ électrique de 2.7 PV.m^-1. Pour cela, ces chaînes lasers délivrent nécessairement des faisceaux de grands diamètres (jusqu'à 40 cm) et des impulsions très courtes (de l'ordre de la dizaine de femtosecondes). En conséquence, les propriétés spatiales et temporelles de l'impulsion ne sont généralement pas indépendantes. Ce type de dépendance, appelée couplage spatio-temporel, a pour conséquence d'augmenter la durée d'impulsion et la taille de la tache focale, ce qui peut conduire à une diminution notable de l'intensité maximale au foyer. Les dispositifs de métrologie couramment utilisés sur ces chaînes lasers femtosecondes de haute puissance ne permettent de mesurer les profils spatial et temporel de l'impulsion que de façon indépendante.L'objectif de cette thèse était de développer des techniques permettant de mesurer les couplages spatio-temporels afin de pouvoir quantifier leur effet et de les corriger dans l'optique d'obtenir l'intensité maximale au foyer. Ainsi, nous avons tout d'abord adapté une technique de caractérisation spatio-temporelle existante à la mesure de lasers TW. Afin d'éviter les contraintes induites au foyer, comme celles liées aux fluctuations de pointé, les mesures ont été réalisées sur le faisceau collimaté. Ajouter une source de référence en parallèle du dispositif initial, nous a aussi permis de prendre en compte les artéfacts de mesure dus aux variations thermiques et mécaniques affectant l'interféromètre. Grâce à cette amélioration, il est possible de reconstruire le profil spatio-temporel complet du faisceau, en particulier son front d'onde.Cependant, les limitations induites par cette technique, nous ont conduit à développer un nouveau dispositif de mesure. Basé sur une corrélation croisée, cette technique consiste à faire interférer le faisceau laser à caractériser avec une partie de ce dernier, suffisamment petite pour ne pas être distordue spatio-temporellement. Nous avons également mis en œuvre une variante de ce dispositif permettant une mesure mono-coup selon une dimension transverse de l'impulsion.A l'aide de ces différentes techniques, nous avons pu caractériser, pour la première fois, plusieurs chaînes lasers TW. Les mesures réalisées ont mis en lumière l'existence de couplages spatio-temporels résiduels conduisant à une baisse significative de l'intensité pic au foyer. Ces résultats montrent qu'il est indispensable de caractériser spatio-temporellement des chaînes lasers femtosecondes de haute puissance dans l'optique d'obtenir l'intensité maximale au foyer. / One of the advantages of high-power femtosecond lasers (TW-PW) is to obtain, at the focus of a focusing optic, very high intensities up to 10^22W.cm^-2 (i.e. an electric field of 2.7 PV.m^-1. Therefore, these lasers chains necessarily deliver beams with large diameter (up to 40 cm) and very short pulses (of the order of tens of femtoseconds). As a consequence, the spatial and temporal properties of the pulse are generally not independent. Such dependence, called spatial-temporal coupling has the effect of increasing the pulse duration and the size of the focal spot, which can lead to a significant reduction of the maximum intensity at the focus. Metrology devices commonly used on these high-power femtosecond lasers allow retriving the spatial and temporal profiles of the pulse only in an independent manner. The aim of this thesis was to develop techniques for measuring spatiotemporal couplings in order to quantify their effect and correct them in order to obtain the maximum intensity at focus. First of all, we adapted an existing technique of spatio-temporal characterization to the measurement of TW lasers. To avoid the issues induced at the focus, such as those related to jittering, measurements were performed on the collimated beam. By adding a reference source to the original device, we managed to take into account the measurement artifacts due to thermal and mechanical variations affecting the interferometer. With this improvement, it was possible to reconstruct the complete spatio-temporal profile of the beam, particularly its wavefront. However, the limitations imposed by this technique led to the development of a new measurement device. Based on a cross-correlation, this technique consists of making the laser beam to interfere with a part of itself, small enough not to be spatiotemporally distorted. We have also implemented a variant of this device for a single-shot measurement along one transverse dimension of the pulse. Using these techniques, we performed the very first characterization of several TW laser chains. The measurements have revealed the existence of residual space-time couplings leading to a significant decrease of the peak intensity at focus. These results show that it is essential to spatio-temporally characterize high power femtosecond laser chains to obtain the maximum intensity at focus.

O lançamento de esgotos domésticos in natura, efluentes das estações de tratamento de esgoto e escoamento superficial, são relatados como importantes causas de poluição das águas superficiais. Sabe-se que a alteração da qualidade das águas dos rios restringe seus múltiplos usos e contribui para o aumento de doenças de veiculação hídrica, em decorrência da exposição oral a bioagentes patogênicos. Neste contexto, o objetivo do presente estudo foi identificar e quantificar bioagentes presentes na água do rio Pardo, Brasil, e estimar o risco de infecção e de doença por Cryptosporidium spp. e Giardia spp. para a população, devido ao uso do rio como fonte de abastecimento público e recreação de contato primário, por meio da abordagem da Avaliação Quantitativa de Risco Microbiológico (AQRM). Durante os anos de 2015 e 2016, foram realizadas seis coletas de amostras da água do rio Pardo (período chuvoso e período seco) em seis pontos, totalizando 36 amostras. Foram realizadas análises de identificação e quantificação de E. coli, Salmonella Não Tifóide, Cryptosporidium spp. e Giardia spp. Para estimativa de risco de infecção e de doença por Cryptosporidium spp. e Giardia spp. (AQRM), foram considerados diferentes populações (crianças e adultos), volumes de água ingerido, concentração de (oo)cistos e duração e frequência da exposição, de acordo com o cenário estabelecido. Os valores médios para E. coli variaram de 6,57 x 101 UFC/100 mL a 6,07 x 103 UFC/100 mL, apresentando diferenças estatisticamente significantes (p < 0,05) entre os períodos chuvoso e seco. As densidades de Salmonella Não Tifóide foram baixas (<0,6473 a 1,55 NMP/100 mL), com frequência de 13,9% das amostras positivas, evidenciando a circulação desse patógeno no ambiente. A concentração de (oo)cistos de Cryptosporidium spp. e Giardia spp. variou de <0,1 a 0,4 oocistos/L e <0,1 a 4,4 cistos/L, respectivamente. Para abordagem da AQRM devido a ingestão da água do rio Pardo usada para abastecimento público, a probabilidade anual de infecção por Cryptosporidium spp. e Giardia spp. foi maior para adultos do que para crianças, sendo que na maioria dos pontos apresentou resultados superiores ao risco anual tolerável pela USEPA (1 x 10-4). No que diz respeito ao uso da água do rio Pardo para recreação de contato primário, a probabilidade diária e anual de infecção, bem como a probabilidade de doenças, foi maior para crianças, seguida de adultos/homens e adultos/mulheres. A probabilidade de criptosporidiose e giardíase esteve abaixo do limite tolerável pela USEPA (3,6 x 10-2), exceto no Ponto 4, em que a estimativa de risco de doença por Giardia spp. para crianças esteve acima deste valor. A presença de bioagentes em amostras de água do rio Pardo pode estar relacionada à poluição das águas por fontes pontuais e difusas. Esses achados refletem a importância de priorizar os recursos para implantação e complementação das Estações de Tratamento de Esgoto na UGRHI 4, a fim de prevenir as doenças de veiculação hídrica em populações que utilizam a água do rio Pardo para abastecimento público e recreação de contato primário / The discharge of domestic sewage, effluents of wastewater treatment plants and surface runoff, are reported as important causes of surface water pollution. It is known that the alteration of river water quality restricts its multiple uses and contributes to the increase of waterborne diseases, due to oral exposure to pathogenic bioagents. In this context, the aim of the present study was to identify and quantify bioagents present in Pardo river water, Brazil, and to estimate the risk of infection and disease by Cryptosporidium spp. and Giardia spp. for the population, due to the use of the river as source of public supply and primary contact recreation, through the approach of Quantitative Microbial Risk Assessment (QMRA). During the years of 2015 and 2016, six samples of water from the Pardo river (rainy and dry season) were collected at six points, totaling 36 samples. Identification and quantification analyzes of E. coli, Non-typhoid Salmonella, Cryptosporidium spp. and Giardia spp. To estimate the risk of infection and disease by Cryptosporidium spp. and Giardia spp. (QMRA), different populations (children and adults), volumes of ingested water, concentration of (oo) cysts, duration and frequency of exposure were considered according to the established scenario. Mean values for E. coli varied from 6.57 x 101 CFU / 100 mL to 6.07 x 103 CFU / 100 mL, showing statistically significant differences (p <0.05) between the rainy and dry season. Non-typhoid Salmonella densities were low (<0.6473 at 1.55 MPN / 100 mL), with a frequency of 13.9% of the positive samples, evidencing the circulation of this pathogen in the environment. Cryptosporidium spp. and Giardia spp. concentration ranged from <0.1 to 0.4 oocysts / L and <0.1 to 4.4 cysts / L, respectively. In order to approach the QMRA due to the ingestion of Pardo river water used for public supply, the probability of annual infection by Cryptosporidium spp. and Giardia spp. was higher for adults than for children, and in most points presented results higher than the risk tolerable by USEPA (1 x 10-4). Regarding the use of Pardo river water for primary contact recreation, the daily and annual probability of infection, as well as the probability of illness, was higher for children, followed by adults / men and adults / women. The probability of cryptosporidiosis and giardiasis was below the limit tolerable by USEPA (3.6 x 10-2), except in Point 4, where the estimated risk of disease by Giardia spp. for children was above this value. The presence of bioagents in Pardo river water may be related to water pollution by point and diffuse sources. These findings reflect the importance of prioritizing the resources for implementation and complementation of wastewater treatment plants at UGRHI 4, in order to prevent waterborne diseases in populations that use Pardo river water for public supply and primary contact recreation

Wireless communications have been an indispensable aspect of everyday life, and there is an increasing consumer demand for accessing several wireless communication technologies from a single, compact, mobile device. System-on-package (SOP) technology is an advanced packaging technology that has been proven to realize the convergence of multiple functions into miniaturized, high-performance systems to meet this demand. With the advancements in the SOP technology, the miniaturization of the front-end module has been achieved using embedded passives in multilayer packages. However, the integration of the antenna directly on the module package is still the barrier to achieve a fully-integrated, high-performance RF SOP system. The main reason for this missing link is that integrating the antenna on the package requires miniaturizing the antenna, which is a difficult task. The focus of this dissertation is to design high-performance antennas along with developing techniques for miniaturization and system-on-package (SOP) integration of these antennas to achieve fully-integrated SOP systems using advanced multilayer organic substrates and thin-film magneto-dielectric materials. The targeted spectrum for the antenna designs are 2.4/5 GHz WLAN/WiMAX and 60 GHz WPAN bands. Several novel antenna designs and configurations to integrate the antenna on the package along with the module are discussed in this dissertation. The advanced polymers used in this research are Liquid Crystalline Polymer (LCP), RXP, and thin-film magneto-dielectrics.

O lançamento de esgotos domésticos in natura, efluentes das estações de tratamento de esgoto e escoamento superficial, são relatados como importantes causas de poluição das águas superficiais. Sabe-se que a alteração da qualidade das águas dos rios restringe seus múltiplos usos e contribui para o aumento de doenças de veiculação hídrica, em decorrência da exposição oral a bioagentes patogênicos. Neste contexto, o objetivo do presente estudo foi identificar e quantificar bioagentes presentes na água do rio Pardo, Brasil, e estimar o risco de infecção e de doença por Cryptosporidium spp. e Giardia spp. para a população, devido ao uso do rio como fonte de abastecimento público e recreação de contato primário, por meio da abordagem da Avaliação Quantitativa de Risco Microbiológico (AQRM). Durante os anos de 2015 e 2016, foram realizadas seis coletas de amostras da água do rio Pardo (período chuvoso e período seco) em seis pontos, totalizando 36 amostras. Foram realizadas análises de identificação e quantificação de E. coli, Salmonella Não Tifóide, Cryptosporidium spp. e Giardia spp. Para estimativa de risco de infecção e de doença por Cryptosporidium spp. e Giardia spp. (AQRM), foram considerados diferentes populações (crianças e adultos), volumes de água ingerido, concentração de (oo)cistos e duração e frequência da exposição, de acordo com o cenário estabelecido. Os valores médios para E. coli variaram de 6,57 x 101 UFC/100 mL a 6,07 x 103 UFC/100 mL, apresentando diferenças estatisticamente significantes (p < 0,05) entre os períodos chuvoso e seco. As densidades de Salmonella Não Tifóide foram baixas (<0,6473 a 1,55 NMP/100 mL), com frequência de 13,9% das amostras positivas, evidenciando a circulação desse patógeno no ambiente. A concentração de (oo)cistos de Cryptosporidium spp. e Giardia spp. variou de <0,1 a 0,4 oocistos/L e <0,1 a 4,4 cistos/L, respectivamente. Para abordagem da AQRM devido a ingestão da água do rio Pardo usada para abastecimento público, a probabilidade anual de infecção por Cryptosporidium spp. e Giardia spp. foi maior para adultos do que para crianças, sendo que na maioria dos pontos apresentou resultados superiores ao risco anual tolerável pela USEPA (1 x 10-4). No que diz respeito ao uso da água do rio Pardo para recreação de contato primário, a probabilidade diária e anual de infecção, bem como a probabilidade de doenças, foi maior para crianças, seguida de adultos/homens e adultos/mulheres. A probabilidade de criptosporidiose e giardíase esteve abaixo do limite tolerável pela USEPA (3,6 x 10-2), exceto no Ponto 4, em que a estimativa de risco de doença por Giardia spp. para crianças esteve acima deste valor. A presença de bioagentes em amostras de água do rio Pardo pode estar relacionada à poluição das águas por fontes pontuais e difusas. Esses achados refletem a importância de priorizar os recursos para implantação e complementação das Estações de Tratamento de Esgoto na UGRHI 4, a fim de prevenir as doenças de veiculação hídrica em populações que utilizam a água do rio Pardo para abastecimento público e recreação de contato primário / The discharge of domestic sewage, effluents of wastewater treatment plants and surface runoff, are reported as important causes of surface water pollution. It is known that the alteration of river water quality restricts its multiple uses and contributes to the increase of waterborne diseases, due to oral exposure to pathogenic bioagents. In this context, the aim of the present study was to identify and quantify bioagents present in Pardo river water, Brazil, and to estimate the risk of infection and disease by Cryptosporidium spp. and Giardia spp. for the population, due to the use of the river as source of public supply and primary contact recreation, through the approach of Quantitative Microbial Risk Assessment (QMRA). During the years of 2015 and 2016, six samples of water from the Pardo river (rainy and dry season) were collected at six points, totaling 36 samples. Identification and quantification analyzes of E. coli, Non-typhoid Salmonella, Cryptosporidium spp. and Giardia spp. To estimate the risk of infection and disease by Cryptosporidium spp. and Giardia spp. (QMRA), different populations (children and adults), volumes of ingested water, concentration of (oo) cysts, duration and frequency of exposure were considered according to the established scenario. Mean values for E. coli varied from 6.57 x 101 CFU / 100 mL to 6.07 x 103 CFU / 100 mL, showing statistically significant differences (p <0.05) between the rainy and dry season. Non-typhoid Salmonella densities were low (<0.6473 at 1.55 MPN / 100 mL), with a frequency of 13.9% of the positive samples, evidencing the circulation of this pathogen in the environment. Cryptosporidium spp. and Giardia spp. concentration ranged from <0.1 to 0.4 oocysts / L and <0.1 to 4.4 cysts / L, respectively. In order to approach the QMRA due to the ingestion of Pardo river water used for public supply, the probability of annual infection by Cryptosporidium spp. and Giardia spp. was higher for adults than for children, and in most points presented results higher than the risk tolerable by USEPA (1 x 10-4). Regarding the use of Pardo river water for primary contact recreation, the daily and annual probability of infection, as well as the probability of illness, was higher for children, followed by adults / men and adults / women. The probability of cryptosporidiosis and giardiasis was below the limit tolerable by USEPA (3.6 x 10-2), except in Point 4, where the estimated risk of disease by Giardia spp. for children was above this value. The presence of bioagents in Pardo river water may be related to water pollution by point and diffuse sources. These findings reflect the importance of prioritizing the resources for implementation and complementation of wastewater treatment plants at UGRHI 4, in order to prevent waterborne diseases in populations that use Pardo river water for public supply and primary contact recreation

<p>    </p> <p>In order to advance next-generation communication systems, it is critical to enhance the state-of-the-art communication architectures, such as device-to-device (D2D), multiple- input multiple-output (MIMO), and intelligent reflecting surface (IRS), in terms of achieving high data rate, low latency, and high energy efficiency. In the first part of this dissertation, we address joint learning and optimization methodologies on cutting-edge network archi- tectures. First, we consider D2D networks equipped with MIMO systems. In particular, we address the problem of minimizing the network overhead in D2D networks, defined as the sum of time and energy required for processing tasks at devices, through the design for MIMO beamforming and communication/computation resource allocation. Second, we address IRS-assisted communication systems. Specifically, we study an adaptive IRS control scheme considering realistic IRS reflection behavior and channel environments, and propose a novel adaptive codebook-based limited feedback protocol and learning-based solutions for codebook updates. </p> <p><br></p> <p>Furthermore, in order for revolutionary innovations to emerge for future generations of communications, it is crucial to explore and address fundamental, long-standing open problems for communications, such as the design of practical codes for a variety of important channel models. In the later part of this dissertation, we study the design of practical codes for feedback-enabled communication channels, i.e., feedback codes. The existing feedback codes, which have been developed over the past six decades, have been demonstrated to be vulnerable to high forward/feedback noises, due to the non-triviality of the design of feedback codes. We propose a novel recurrent neural network (RNN) autoencoder-based architecture to mitigate the susceptibility to high channel noises by incorporating domain knowledge into the design of the deep learning architecture. Using this architecture, we suggest a new class of non-linear feedback codes that increase robustness to forward/feedback noise in additive White Gaussian noise (AWGN) channels with feedback. </p>