The design of power amplifiers within a circuit simulator requires a good non-linear model that accurately predicts the electormagnetic behaviour of the power transistor. In recent years, a certain class of large signal frequency-dependent black-box behavioural modeling techniques known as Poly-Harmonic Distortion (PHD) models has been devised to mimic the non-linear unmatched RF transistor. These models promise a good prediction of the device behaviour under multi-harmonic periodic continuous wave inputs. This thesis describes the capabilities of the PHD modeling framework and the theoretical type of behaviour that it is capable of predicting. Specifically, the PHD framework cannot necessarily predict the response of a broadband aperiodic signal. This analysis will be performed by deriving the PHD modeling framework as a simplification of the Volterra series kernel functions under the assumption that the power transistor is operating under continuous periodic multi-harmonic voltage and current signals in a stable circuit. A PHD model will be seen as a set of describing functions that predict the response of the Device Under Test (DUT) for any given non-linear periodic continuous-wave inputs that have a specific fundamental frequency. Two popular implementations of PHD models that can be found in the literature are the X-parameter and Cardiff models. Each model formulates the describing functions of the general PHD model differently. The mathematical formulation of the X-parameter and Cardiff models will be discussed in order to provide a theoretical ground for comparing their robustness. The X-parameter model will be seen as the first-order Taylor series approximation of the PHD model describing functions around a Large Signal Operating Point (LSOP) of the device under test. The Cardiff large-signal model uses Fourier series coefficient functions that vary with the magnitude of the large signal(s) as the PHD model describing functions. This thesis will provide a breakdown of the measurement procedure required for the extraction of these models, the challenges involved in the measurement, as well as the mathematical extraction of the model coe cients from measurement data. As each of these models contain have extended versions that enhance the predictive capability of the model under stronger nonlinear modes of operation, a comparison is used to represent the cost of increasing model accuracy as a function of the increasing model complexity for each model. The order of complexity of each model can manifest itself in terms of the mathematical formulation, the number of parameters required and the measurement time that is required to extract each model for a given DUT. This comparison will fairly assess the relative strengths and weaknesses of each model.

The design of power amplifiers within a circuit simulator requires a good non-linear model that accurately predicts the electormagnetic behaviour of the power transistor. In recent years, a certain class of large signal frequency-dependent black-box behavioural modeling techniques known as Poly-Harmonic Distortion (PHD) models has been devised to mimic the non-linear unmatched RF transistor. These models promise a good prediction of the device behaviour under multi-harmonic periodic continuous wave inputs. This thesis describes the capabilities of the PHD modeling framework and the theoretical type of behaviour that it is capable of predicting. Specifically, the PHD framework cannot necessarily predict the response of a broadband aperiodic signal. This analysis will be performed by deriving the PHD modeling framework as a simplification of the Volterra series kernel functions under the assumption that the power transistor is operating under continuous periodic multi-harmonic voltage and current signals in a stable circuit. A PHD model will be seen as a set of describing functions that predict the response of the Device Under Test (DUT) for any given non-linear periodic continuous-wave inputs that have a specific fundamental frequency. Two popular implementations of PHD models that can be found in the literature are the X-parameter and Cardiff models. Each model formulates the describing functions of the general PHD model differently. The mathematical formulation of the X-parameter and Cardiff models will be discussed in order to provide a theoretical ground for comparing their robustness. The X-parameter model will be seen as the first-order Taylor series approximation of the PHD model describing functions around a Large Signal Operating Point (LSOP) of the device under test. The Cardiff large-signal model uses Fourier series coefficient functions that vary with the magnitude of the large signal(s) as the PHD model describing functions. This thesis will provide a breakdown of the measurement procedure required for the extraction of these models, the challenges involved in the measurement, as well as the mathematical extraction of the model coe cients from measurement data. As each of these models contain have extended versions that enhance the predictive capability of the model under stronger nonlinear modes of operation, a comparison is used to represent the cost of increasing model accuracy as a function of the increasing model complexity for each model. The order of complexity of each model can manifest itself in terms of the mathematical formulation, the number of parameters required and the measurement time that is required to extract each model for a given DUT. This comparison will fairly assess the relative strengths and weaknesses of each model.

Level-set methods have become a valuable and well-established field of visualization over the last decades. Different implementations addressing different design goals and different data types exist. In particular, level sets can be used to extract isosurfaces from scalar volume data that fulfill certain smoothness criteria. Recently, such an approach has been generalized to operate on unstructured point-based volume data, where data points are not arranged on a regular grid nor are they connected in form of a mesh. Utilizing this new development, one can avoid an interpolation to a regular grid which inevitably introduces interpolation errors. However, the global processing of the level-set function can be slow when dealing with unstructured point-based volume data sets containing several million data points. We propose an improved level-set approach that performs the process of the level-set function locally. As for isosurface extraction we are only interested in the zero level set, values are only updated in regions close to the zero level set. In each iteration of the level-set process, the zero level set is extracted using direct isosurface extraction from unstructured point-based volume data and a narrow band around the zero level set is constructed. The band consists of two parts: an inner and an outer band. The inner band contains all data points within a small area around the zero level set. These points are updated when executing the level set step. The outer band encloses the inner band providing all those neighbors of the points of the inner band that are necessary to approximate gradients and mean curvature. Neighborhood information is obtained using an efficient kd-tree scheme, gradients and mean curvature are estimated using a four-dimensional least-squares fitting approach. Comparing ourselves to the global approach, we demonstrate that this local level-set approach for unstructured point-based volume data achieves a significant speed-up of one order of magnitude for data sets in the range of several million data points with equivalent quality and robustness.

Optical imaging methods incorporating observation of the mucosal surface of organs with narrowband filtered light are diagnostic tools with a significant development in recent years and have become an integral and inseparable part of the diagnostic process of benign and malignant pathologies of the mucosal surface of various anatomical systems of the human body. Malignant and premalignant lesions of the upper aero-digestive tract (UADT) are no exception. The present work demonstrates the benefits of evaluation of changes in microvascular architecture by narrowband imaging. Its major advantage is a more accurate diagnosis of premalignant and malignant mucosal pathologies in the head and neck region based on the identification of significant changes in superficial vascularity. There is a very strong correlation of these vascular changes with the final histopathological analysis. The present data have shown that videoendoscopic examination with narrowband imaging significantly improves the distinction between benign and malignant mucosal lesions even in the UADT pathologies with a similar tendency to neoangiogenesis (squamous cell papilloma vs. squamous cell carcinoma). It also refines the determination of surface extension of tumoral changes and helps to localize and accurately diagnose tumors of...

Diagnosis of laryngeal and hypopharyngeal cancer is often in advanced stages of cancer, resulting in elimination of the possibility of treatment with the least mutilating effect for the patient. The huge progamination. ress in diagnostic process of mucosal tumors of the head and neck has been started with the development of new optical endoscopic imaging methods that are able to visualize initiating discrete mucosal changes with pathological vascularization easier and more accurately. The aim of this dissertation is to evaluate the benefit of Narrow Band Imaging (NBI), especially in the differentation of benign, precancerous and malignant changes in three groups of patients, it means in preoperative diagnostics in outpatient department (group I), during the endoscopic surgery (group II) and also in follow-up of patients after curative radiotherapy (group III). The vascular changes within the observed lesion or in close contact with the laryngeal/hypopharyngeal lesion were classified according to two classifications (Ni and Arens) and subsequently correlated with the histopathological results. The dissertation confirmed that NBI plays a key role in early detection of precancerous and malignant changes, in discrimination between benign and malignant vascular patterns with statistical significance. We...

No description available.

L’engouement actuel pour les applications de type réseaux de capteurs sans-fil ou internet des objets (IoT) relance la nécessité, alors initiée par les applications mobiles, de concevoir des émetteurs-récepteurs radio à basse consommation. Dans ce contexte, l’objet des travaux de thèse est de proposer des techniques de réduction de la consommation des récepteurs radio tout en minimisant l’impact sur leur architecture de manière à pouvoir adapter leur consommation aux besoins de performance.Pour ce faire, l’utilisation intermittente du convertisseur analogique numérique (ADC) a, dans un premier temps, été étudiée puis celle-ci a été généralisée à l’ensemble du récepteur. Pour chacune de ces approches, une modélisation de la dégradation des performances en termes de taux d’erreur (BER) a été confrontée à une estimation de la réduction de la consommation engendrée. Par ailleurs, l’impact de l’ajout de modules spécifiques aux techniques proposées est décrit à l’aide de résultats concernant leurs complexités et leurs consommations. L’ensemble de ces résultats s’inscrit pleinement dans le domaine de recherche des récepteurs adaptatifs pour lesquels les performances sont adaptées au canal de transmission en temps réel.Finalement, une technique de compensation digitale des défauts de quadrature a été proposée, rendant possible l’utilisation d’une PLL moins énergivore mais avec des performances dégradées. Cette technique utilise une recherche par dichotomie des poids de compensation des défauts de quadrature, lui permettant de converger suffisamment rapidement pour pouvoir réaliser la compensation sur une portion connue du message reçu et ainsi éviter une perte d’information. / The recent craze for the Wireless Sensor Networks (WSN) and the Internet of Things (IoT) applications boosts the necessity, previously introduced by the mobile applications, to design low power transceivers. In this context, the purpose of this thesis is to propose some techniques to reduce the power consumption of RF receivers while minimizing the impact on their architecture in order to be able to adapt their power consumption to the required performances.To do so, the study of the intermittent use of the analog-to-digital converter (ADC) is firstly proposed and then extended to the whole receiver. In each case, the degradation of the receiver performances in terms of bit error rate (BER) is compared to an estimate of the obtained decrease of the power consumption. Moreover, the complexity and the overhead power consumption of the modules involved in the processing of the proposed techniques are also estimated and discussed. All these results are part of the field of research called “adaptive receiver” that tries to adapt the receiver performances to its environment in real time.Finally, a digital compensation technique of the quadrature imbalances was proposed. It allows using a less energy-consuming PLL but with degraded quadrature performances and compensating the mismatches in the digital domain. This technique uses a dichotomic search of the compensation weights allowing a fast convergence in order for the compensation to be done during the reception of a known portion of the received message and therefore avoiding a loss of information.

No description available.

La compagnie Sigfox est reconnue comme un acteur prometteur pour des transmissions de longue-distance et faible consommation, dans le contexte de l'IoT. La modulation à bande ultra étroite (Ultra Narrow Band (UNB)), la technologie de communication choisie par Sigfox, permet de transmettre des informations dans des bandes de signal très étroites (typiquement 100 Hz). A cause de l'imprécision fréquentielle causée par les oscillateurs générateurs de fréquence, il n'est pas réaliste de transmettre des signaux UNB dans des canaux parfaitement orthogonaux. L'accès naturel au canal radio pour le système de UNB est de type ALOHA, avec un aspect aléatoire à la fois en en temps et en fréquence. Cet accès aléatoire peut introduire des collisions qui dégradent la performance du réseau. Le but de cette thèse est de caractériser la capacité des réseaux basés sur UNB, ainsi que d’améliorer la performance en considérant l'aspect aléatoire en temps et en fréquence. La première contribution de cette thèse, est une évaluation de la capacité en théorie et en simulation pour une seule station de base (BS), sous des conditions de canal idéaliste ou réaliste. En conditions idéalistes, nous avons exprimé la capacité pour le cas de l'ALOHA généralisé, et l'avons étendu aux cas de réplications. Pour les conditions réalistes, nous avons pris en compte l'interférence spectrale d'UNB et le path loss (sans et avec Rayleigh fading) afin de caractériser la performance des réseaux UNB, avec l'outil géométrie stochastique. La deuxième contribution est d'appliquer l’annulation successive d'interférence (SIC), qui nous permet d'atténuer les interférences, dans des réseaux de UNB. Nous avons fourni une analyse théorique de la performance des réseaux en considérant le SIC et l'interférence spectrale de UNB, pour le cas de mono-BS. La troisième contribution est l'amélioration de la performance des réseaux UNB, en exploitant la diversité de multi-BS. Nous avons fait une analyse théorique de performance en considérant multi-BS et selection combining (SC). En particulier, nous avons considéré que l’interférence vue par chaque BS est corrélée. Nous avons ainsi démontré mathématiquement que cette corrélation ne peut pas être supprimée dans des systèmes UNB. Ensuite, nous avons appliqué les technologies de la combinaison des signaux plus complexes comme MRC (max ratio combining) et EGC (equal gain combining), ainsi que le SIC à travers multi-BS. Nous avons évalué l'amélioration de performance que chaque technologie apporte, et les avons comparées. Nous avons souligné l'efficacité de ces technologies qui nous permettent d’obtenir des gains importants comparés au cas mono-BS (e.x. 125 fois plus de réduction d'erreur avec SIC globale). La dernière contribution est une validation expérimentale du modèle d'interférence spectrale de UNB, ainsi que la capacité des réseaux UNB, sur un testbed de radio FIT/Cortexlab. / Sigfox rises as a promising candidate dedicated for long-distance and low-power transmissions in the IoT backgrounds. Ultra Narrow Band (UNB), being the communication technology chosen by Sigfox, allows to transmit information through signals whose bandwidth is very limited, typically 100 Hz. Due to the imprecision restraint on electronic devices, it is impossible to transmit UNB signals in orthogonal channels. The natural radio access for this kind of system is thus random ALOHA, in both time and frequency domain. This random access can induce collisions which degrades the networks performance. The aim of this thesis is to characterize the capacity of UNB based networks, as well as to enhance its performance, by considering the randomness in time and frequency. The first contribution of the thesis, is the theoretical and numerical capacity evaluation under idealized and realistic channel conditions, for mono base station (BS) case. Under idealized conditions, we have quantified this capacity for generalized ALOHA case and extended for replications. We highlight the time-frequency duality in UNB systems, and that there exists an optimum replication number for a given network parameter set. Under realistic conditions, we have taken into account the specific spectral interference of UNB systems and propagation path loss (without and with Rayleigh fading) to characterize the performance, with the aid of stochastic geometry. The second contribution is the enhancement of UNB network performance in single BS case. We propose to use successive interference cancellation (SIC) in UNB networks, which allows to mitigate the interference. We have provided a theoretical analysis by considering both SIC and the spectral interference, for mono-BS case. We bring to light the efficiency of SIC in enhancing UNB system performance. The third contribution is the improvement of UNB systems, by exploiting the multiple BS diversity. An analytical performance evaluation considering the simplest selection combining is conducted. In particular, we consider the interference viewed by all the BSs are correlated. Then we apply more complex signal combining technologies such as MRC (max ratio combining) and EGC (equal gain combining), and even interference cancellation across multi-BS in UNB networks. We evaluate the performance improvement that each technology can bring, and compare them with each other. We highlight the efficiency of these multi-BS technologies which allow us to achieve significant performance enhancement compared to mono-BS (e.x. 125 times better performance with global SIC). Last but not least, we experimentally verify the the spectral interference model and network capacity on a cognitive radio testbed.

Cette thèse a pour but de modéliser les réseaux de capteurs sans fil à faible débit (WSN) basés sur la technique ultra bande étroite. Ce réseau a déjà été déployée par la société SIGFOX et a déjà démontré sa très grande efficacité pour les applications pour l’Internet des objets (IoTs) grâce à sa capacité de communication point à point efficace en terme de puissance consommée, et de sa connectivité de longue portée. Cette étude donne quelques aperçus sur le passage à l’échelle de la technique de l’UNB pour un réseau multipoint à point pour une liaison montante. L’accès au canal spécifique qui est basé sur l’accès multiple par répartition aléatoire de fréquence et de temps (R-FTDMA) est introduit et analysé. En outre, l’impact de l’interférence due à l’absence de stratégie d’ordonnancement à la couche MAC est étudié et modélisé. Notre modèle simplifié nous permet non seulement de décrire la puissance d’interférence agrégée, mais aussi d’évaluer les performances du système d’un tel réseau en matière de taux d’erreur et de probabilité de coupure. De même, la géométrie stochastique est utilisée pour modéliser la distribution spatiale des noeuds afin d’étendre le modèle simplifié dans le canal réaliste où les dégradations de canal sont prises en compte. De plus, le mécanisme de retransmission est considéré pour ce réseau. Cette étude permet de d’identifier le nombre optimal de retransmissions. Le réseau peut être configuré avec un paramètre global unique. Et enfin, cette thèse met en évidence le fait que le réseau de l’UNB Random-FTDMA est très pertinent dans un réseau réaliste, en particulier pour les applications à faible débit, car il allége le coût élevé du réseau, le coût de la synchronisation globale, mais sans perte de performance. / This thesis aims at modeling the low-throughput wireless sensor networks (WSNs) based on ultra-narrow-band technology. Such wireless network is already been deployed by Sigfox company and has proved to be ultra-efficient for the Internet of things (IoTs) applications thanks to its ability of point-to-point communication in terms of power efficiency and long range connectivity. In particular, this thesis gives some insights on the scalability of UNB technology for a multi-point-to-point network in an uplink scenario. The multiple access schemes based on random time and frequency selection are introduced and analyzed. Furthermore, the interference impact due to the lack of scheduling strategy at the MAC layer is studied and modeled. Our simplified model using rectangular function allows us not only to describe the aggregate interference power but also evaluate the system performance of such network in terms of the bit-error-rate and outage probability. Besides, the geometry stochastic is used for spatial node distribution in order to extend the simplified model in the realistic channel communication where the channel impairments are taken into account. Besides, the retransmission mechanism is considered for such network. This study argues an optimal number of retransmission. The network can be configured with a unique global parameter. Last but not least, this thesis highlights the fact that the UNB network using Random-FTDMA schemes is very relevant in a realistic network, especially for low-throughput applications, because it bypasses the high network cost, the cost of global synchronization but without loss of performance.