Approved for public release; distribution in unlimited. / While a necessity of all operating systems, the code that initializes a system can be notoriously difficult to understand. This thesis explores the most common architectures used for bringing an operating system to its initial state, once the operating system gains control from the boot loader. Specifically, the ways in which the OpenBSD and Linux operating systems handle initialization are dissected. With this understanding, a set of threats relevant to the initialization sequence was developed. A thorough study was also made to determine the degree to which initialization code adheres to widely accepted software engineering principles. Based upon this threat analysis and the observed strengths and weaknesses of existing systems, a set of recommendations for initialization sequence architecture and implementation have been developed. These recommendations can serve as a guide for future operating system development. / Civilian, Naval Postgraduate School

Durant ma thèse, mes recherches se sont principalement portées sur l'identification des propriétés modulaires de la dysferline, une protéine impliquée dans des dystrophies musculaires, afin d'identifier les approches thérapeutiques les plus prometteuses. Mon travail s’est donc orienté selon deux axes de recherche, une approche par « mini-protéines » et une approche par «saut d’exon », toutes deux basées sur des preuves de principe obtenues chez des patients. Nous avons tout d'abord testé une approche de saut d'exon pour l'exon 32 de DYSF. Nous avons pu établir la fonctionnalité de cette protéine tronquée en permettant son expression dans des cellules de patients déficients en dysferline. Ceci a suggéré que le domaine C2D (encodé par les exons 31 à 34) n'est pas essentiel pour la dysferline puisque l'absence d'une partie de celui-ci n'empêche pas sa fonctionnalité. Nous avons dans le même temps analysé les caractéristiques d'autres domaines de la dysferline, en créant des miniprotéines contenant différentes combinaisons de domaines. Nous avons montré que la partie C-terminale de la dysferline (composée des deux derniers domaines C2 et du passage transmembranaire) était essentielle et suffisante pour la fonctionnalité de la dysferline dans le muscle. L'ensemble de ces résultats démontre que certains domaines de la dysferline sont dispensables, ouvrant ainsi la voie à l'étude d'approches de thérapie génique basé sur les minigènes ou le saut d'exon pour les dysferlinopathies. / During my thesis, my researches have mainly concerned the modular properties of dysferlin, a protein involved in muscular dystrophies, in order to identify the most promising therapeutic approaches.My work has been oriented within two research axes, a mini-proteins approach and an exon-skipping approach, both based on proofs of concept obtained in patients. We first tested an exon-skipping approach for the exon 32 of DYSF, based on the identification of a protein lacking the encoded part of this exon, in an asymptomatic person. We have established the functionality of this truncated protein by allowing its expression in dysferlin-deficient patients' cells. This suggests that the C2D domain (encoded by exons 31 to 34) is not essential for dysferlin since the absence of a part of it don't block its functionality.In the same time we analyzed the characteristics of the others domains of dysferlin, by creating miniproteins containing different combinations of domains. By studying the abilities of this constructs, we have showed that the C-ter part of dysferlin (composed of the last C2 domains and the transmembrane domain) was essential and sufficient for the functionality of dysferlin in muscle. All these results demonstrate that several domains of dysferlin are dispensable, paving the way for studying gene therapy approaches, based on minigenes or exon-skipping for dysferlinopathies.

Cette recherche s’inscrit dans le contexte de la santé en France, dans lequel la prise en charge de la population croissante des usagers atteints de maladies chroniques, nécessite une articulation cohérente des interventions pluridisciplinaires, provenant de multiples d’acteurs hétérogènes. Cette prise en charge personnalisée est très consommatrice de ressources. Dans un contexte de raréfaction des ressources, le dilemme coût-qualité se pose dans leur prise en charge. La recherche d’une réponse à ce dilemme nous a conduit à nous intéresser au concept de la Personnalisation de Masse (PM), largement étudié en Supply Chain Management (SCM) et Management des Opérations (OM), ainsi qu’à sa réalisation dans le contexte de la santé. Selon ces deux domaines de littérature, la réalisation de la PM repose sur le développement de la Capacité de PM (CPM). En outre, l’intégration de la Supply Chain (SCI) et la modularité ont été identifiées comme les deux plus importants facteurs facilitant ce développement. Ainsi, cette recherche vise à étudier le rôle de la SCI et de la modularité, ainsi que leur relation dans le développement de la CPM en santé. En se basant sur les relations étudiées et celles insuffisamment étudiées, entre ces trois concepts en SCM et OM (contexte général), un modèle conceptuel de recherche a été établi, et nous cherchons à éprouver ce modèle dans le contexte de la santé. Pour ce faire, un cas exemplaire a été identifié : la MAIA Marseille. À travers cette étude de cas, notre recherche a permis d’une part d’éprouver et de raffiner le modèle de recherche préétabli, et d’autre part de mettre en avant les leçons à en tirer dans le développement de la CPM en santé. / This research is conducted in the French healthcare context, in which the care of the growing population of health service users with chronic diseases, requires a coherent coordination of multidisciplinary interventions from multiple providers. This personalized care is resource consuming. In a context of resource scarcity, the cost-quality dilemma arises in the care of this growing population. The search for an answer to this dilemma has led us to focus on the concept of Mass Customization (MC), widely studied in Supply Chain Management (SCM) and Operations Management (OM), as well as its implementation in the healthcare context. According to these two fields of literature, MC can be achieved by developing Mass Customization Capability (MCC). Moreover, Supply Chain Integration (SCI) and Modularity-based practices have been identified as the two most important factors facilitating its development. Therefore, this research aims to study the role of SCI and modularity, as well as their relationship in the development of MCC in the healthcare context. Based on the relationships studied and those insufficiently investigated between these three concepts in SCM and OM literature (general context), a conceptual model has been established, and we seek to test this model in the healthcare context. For this purpose, an exemplary/critical case has been identified: MAIA Marseille. Through this case study, our research enabled us to test and refine the pre-established conceptual model, and to highlight the lessons to be learned in the development of MCC in the healthcare context.

Sistemas morfológicos quantitativos são descritos por medidas contínuas. A relação genética entre essas características dos indivíduos é representada pela matriz de covariância genética aditiva, a matriz G. Entender a evolução da matriz G, portanto, é de suma importância para compreender os padrões de diversificação encontrados na natureza. Neste trabalho estudamos modelos computacionais para a evolução de traços contínuos em populações naturais, sujeitas a variados tipos de seleção e condições internas, focando no problema da evolução dos padrões de integração e modularidade nessas populações. Testamos dois modelos com diferentes combinações de parâmetros em sua capacidade de reproduzir e elucidar padrões naturais. Seleção direcional correlacionada se mostrou uma força importante na criação desses padrões de covariação e a seleção estabilizadora correlacionada se mostrou fundamental para a manutenção desses padrões / Quantitative morphological systems are described by continuous measurements. The genetic relation between these characteristics of the individuals is represented by the genetic additive co-variance matrix, the G matrix. Understanding the evolution of the G matrix is, therefore, of paramount importance for proper interpretation of the patterns of diversification we observe in nature. In this work we study computational models for the evolution of quantitative traits in natural populations, subject to different natural selection and internal conditions, focusing on the problem of the evolution of the pattern of morphological integration and modularity. We test two models with different sets of parameters in their ability to reproduce and elucidate natural patterns. Directional correlated selection was necessary for the shaping of the patterns of morphological integration, and correlated stabilizing selection was fundamental to the maintenance of these patterns

O problema de agrupamento de dados em grafos consiste em encontrar clusters de nós em um dado grafo, ou seja, encontrar subgrafos com alta conectividade. Esse problema pode receber outras nomenclaturas, algumas delas são: problema de particionamento de grafos e problema de detecção de comunidades. Para modelar esse problema, existem diversas formulações matemáticas, cada qual com suas vantagens e desvantagens. A maioria dessas formulações tem como desvantagem a necessidade da definição prévia do número de grupos que se deseja obter. Entretanto, esse tipo de informação não está contida em dados para agrupamento, ou seja, em dados não rotulados. Esse foi um dos motivos da popularização nas últimas décadas da medida conhecida como modularidade, que tem sido maximizada para encontrar partições em grafos. Essa formulação, além de não exigir a definição prévia do número de clusters, se destaca pela qualidade das partições que ela fornece. Nesta Tese, metaheurísticas Greedy Randomized Search Procedures para dois modelos existentes para agrupamento em grafos foram propostas: uma para o problema de maximização da modularidade e a outra para o problema de maximização da similaridade intra-cluster. Os resultados obtidos por essas metaheurísticas foram melhores quando comparadas àqueles de outras heurísticas encontradas na literatura. Entretanto, o custo computacional foi alto, principalmente o da metaheurística para o modelo de maximização da modularidade. Com o passar dos anos, estudos revelaram que a formulação que maximiza a modularidade das partições possui algumas limitações. A fim de promover uma alternativa à altura do modelo de maximização da modularidade, esta Tese propõe novas formulações matemáticas de agrupamento em grafos com e sem pesos que visam encontrar partições cujos clusters apresentem alta conectividade. Além disso, as formulações propostas são capazes de prover partições sem a necessidade de definição prévia do número de clusters. Testes com centenas de grafos com pesos comprovaram a eficiência dos modelos propostos. Comparando as partições provenientes de todos os modelos estudados nesta Tese, foram observados melhores resultados em uma das novas formulações propostas, que encontrou partições bastante satisfatórias, superiores às outras existentes, até mesmo para a de maximização de modularidade. Os resultados apresentaram alta correlação com a classificação real dos dados simulados e reais, sendo esses últimos, em sua maioria, de origem biológica / Graph clustering aims at identifying highly connected groups or clusters of nodes of a graph. This problem can assume others nomenclatures, such as: graph partitioning problem and community detection problem. There are many mathematical formulations to model this problem, each one with advantages and disadvantages. Most of these formulations have the disadvantage of requiring the definition of the number of clusters in the final partition. Nevertheless, this type of information is not found in graphs for clustering, i.e., whose data are unlabeled. This is one of the reasons for the popularization in the last decades of the measure known as modularity, which is being maximized to find graph partitions. This formulation does not require the definition of the number of clusters of the partitions to be produced, and produces high quality partitions. In this Thesis, Greedy Randomized Search Procedures metaheuristics for two existing graph clustering mathematical formulations are proposed: one for the maximization of the partition modularity and the other for the maximization of the intra-cluster similarity. The results obtained by these proposed metaheuristics outperformed the results from other heuristics found in the literature. However, their computational cost was high, mainly for the metaheuristic for the maximization of modularity model. Along the years, researches revealed that the formulation that maximizes the modularity of the partitions has some limitations. In order to promote a good alternative for the maximization of the partition modularity model, this Thesis proposed new mathematical formulations for graph clustering for weighted and unweighted graphs, aiming at finding partitions with high connectivity clusters. Furthermore, the proposed formulations are able to provide partitions without a previous definition of the true number of clusters. Computational tests with hundreds of weighted graphs confirmed the efficiency of the proposed models. Comparing the partitions from all studied formulations in this Thesis, it was possible to observe that the proposed formulations presented better results, even better than the maximization of partition modularity. These results are characterized by satisfactory partitions with high correlation with the true classification for the simulated and real data (mostly biological)

L'informatique dans les nuages propose une alternative économique et performante au déploiement traditionnel sur site. Une utilisation efficace des nuages passe par une diminution des ressources consommées et une adaptation du déploiement à l'environnement cible. Nous proposons, dans cette thèse, une solution pour le déploiement de logiciels à la juste taille, c'est-à-dire en ne déployant que les dépendances nécessaires sur un environnement cible. Nous suivons une définition du déploiement basée sur les architectures à composants. Le contenu et les dépendances d'un logiciel sont capturés grâce au modèle à composants Fractal SoftwareUnit. Ce modèle offre une fine granularité de la représentation, le support de l'hétérogénéité des logiciels et des environnement cibles, ainsi qu'un contrôle distribué des logiciels. Nous proposons une mise en œuvre du modèle au travers du canevas SU Framework. Enfin, nous décrivons comment résoudre deux problèmes issus de besoins industriels, à l'aide de ce canevas : la définition de profils pour le serveur d'applications JOnAS et le déploiement d'applications hétérogènes sur cibles hétérogènes. / Cloud computing offers an economical and efficient alternative to traditional deployment on site. Effective use of cloud computing goes through a reduction of resource consumption and deployment tailored to the target environment. We propose, in this thesis, a solution for deploying softwares at fair size, by considering only the necessary dependencies to a target environment. We follow a definition of deployment based on component architecture. The content and dependencies of softwares are captured through the Fractal SoftwareUnit component model. This model offers a fine granularity of representation, the support of software heterogeneity and environmental targets, just as a distributed control of software. SU Framework is an implementation of this model. Finally, we describe how to solve two problems arising from industrial needs, by using the proposed framework : définition of profiles for the JOnAS application server and deployment of heterogeneous applications on heterogeneous targets.

Nas últimas décadas o desenvolvimento de novos produtos se tornou o foco da competitividade industrial e, é neste ambiente que a literatura atual identifica vários casos que apontam para uma forte tendência de descentralização do desenvolvimento de produto em âmbito global. Assim, pretende-se por meio de estudo de caso único, analisar o processo de descentralização de desenvolvimento global de novos produtos (DGP) adotado por uma empresa automotiva atuante no setor de veículos comerciais. O estudo foca na pesquisa dos fatores que favorecem a descentralização do DGP, bem como suas causas, baseado na teoria vigente. Como principais conclusões, pode-se citar que o setor de veículos comerciais reage de maneira similar a outros setores, como o de telecomunicações, e que diferentemente da literatura vigente sobre a descentralização do DGP, fatores como o interesse no mercado local e aspectos históricos da evolução da subsidiária brasileira apresentam maior relevância do que o baixo custo e as competências locais em desenvolvimento de produtos. Identificou-se, também, que neste caso o produto modular influencia no processo de descentralização do DGP dependendo do tipo de estrutura deste último, ao contrário do identificado por pesquisas anteriores. Por fim, é elaborada uma proposta de relação entre tipo de estrutura de DGP e a utilização ou não de projeto modular. / In recent decades the development of new products has become the industrial competitiveness focus and, it is in this environment, that the current literature identifies several cases which indicate a strong trend of product development decentralization in a global sphere. Thus, this study intent to, through a single case study, analyzes the decentralization process of new products global development (PGD) used by a company from the automotive industry acting at the commercial vehicles sector. The study focuses on the research of the factors that encourage the decentralization of PGD and its causes, based on existing theory. As main conclusions, it is possible to mention that the commercial vehicles sector reacts similarly to other sectors such as telecommunications, in which factors such as interest in the local market and historical aspects of evolution of the Brazilian subsidiary have more relevance than the low cost and local skills in product development, unlike the current literature on this topic. It was also identified, in this case, that the modular product influences the PGD decentralization process depending on its structure, unlike identified by previous studies. Finally, the author developed a proposal which describes the relation between the PGD type of structure and the use (or not) of modular design.

Dans le contexte économique actuel, il faut proposer des produits personnalisés dequalité, à faible coût et dans des délais de plus en plus courts. La société MABI a choisi devoir chacune de ces contraintes comme une opportunité de repenser ses produits en misantsur l’innovation. Il faut alors optimiser certaines tâches routinières d’ingénierie afin dedégager du temps pour la conception des nouveaux produits. Le travail de recherche réalisés’inscrit dans le cadre d’une thèse en convention CIFRE en partenariat entre la société MABIet le laboratoire IRTES-M3M de l’UTBM. MABI conçoit, assemble, commercialise et assurele service après-vente de produits propres dans le domaine de la protection et la rénovationdes bâtiments. Ses besoins d’amélioration concernent le processus de développement deproduits qui doivent répondre aux besoins des clients tout en respectant des contraintesd’assemblage spécifiques à l’entreprise. La finalité industrielle de la thèse consiste à décliner au niveau du domaine du« Produit », la méthodologie générique élaborée sur la base de notre travail de recherchescientifique. A ce niveau, notre problématique scientifique consiste à rendre opérationnel etdynamique le modèle Multi-Domaines et Multi-Vues (MD-MV), structuré de manière «plutôtstatique», en y apportant des éléments de raisonnement contribuant à créer des interactionsinter-domaines et inter-points de vue. Pour ce qui est du domaine du « Produit », il endécoule la méthodologie FARD (Functional And Robust Design) qui vise à concevoir et àgénérer rapidement l’ensemble des variantes de produits d’une même famille modulaire touten assurant le respect des besoins clients (conception fonctionnelle) et des contraintesd’assemblage à travers une aide à la décision pour le choix de la séquence d’assemblage,contribuant ainsi à créer une interaction dynamique avec le domaine du « Process ». Quatrethèmes de recherche sont abordés : la modularité, la conception fonctionnelle, la conceptionpour l’assemblage (dès les phases amont du processus de conception) et la simulation(accélérée grâce au paramétrage du maillage). Habituellement, le domaine de la modularitéest souvent associé à celui de la conception fonctionnelle ou encore à celui de la conceptionpour l’assemblage, mais rarement les trois ensemble, ce qui constitue la spécificité de nostravaux. Enfin, l’aspect paramétrique de la méthodologie FARD, à travers les liens établisentre les quatre thèmes de recherche évoqués précédemment, rend possible la générationrapide des produits d’une même famille à partir d’un produit générique et ainsi de gagner dutemps de conception, en vue d‘atteindre nos objectifs de conception routinière « HautementProductive ». Trois cas d’études industriels et académiques illustrent l’application et lafaisabilité la méthodologie FARD... / In current economic context, enterprises must provide quality custom products at alower cost and a shorter delay. MABI Company chose to consider these constraints as anopportunity to rethink its products through innovation. Then certain routine tasks must beoptimized to free up time in order to have more time to innovate and design new products.This thesis is part of a CIFRE partnership between the MABI Company and the IRTES-M3Mlaboratory at UTBM. MABI designs, assembles, sells and provides after-sales service ofproducts in the field of the protection and the renovation of buildings. MABI needs ofimprovement are in the development process of its products that must meet customer needs,while respecting its assembly constraints.The industrial purpose of the thesis is to decline in the Product domain, the genericmethodology developed on the basis of our scientific research work. At this level, ourscientific problematic is to make operational and dynamic the Multi-Domains and Multi-Viewsdesign model (MD-MV), structured in a "rather static" way, and to enriched these models byadding reasoning procedures. It follows the FARD methodology (Functional And RobustDesign) which aims to design and quickly generate variants of a modular product family whileensuring compliance with customer requirements (functional design) and assemblyconstraints. Four domains are covered: modularity, functional design, design for assembly (atthe early stages of the design process) and simulation (accelerated through theparameterisation of the mesh). Usually the domain of modularity is often associated withfunctional design or with the design for assembly, but rarely the three together, thatconstitutes one of our added values. Finally, the parametric aspect of the FARDmethodology, that is the link between the four domains, allows accelerated the generation ofproducts of the same family from a generic product and thus saving design time to achieveour goal of "High Productive" routine design. Three industrial and academic case studiesillustrate the application and the feasibility of the FARD methodology...

The thesis studies Le Corbusier&rsquo / s Dom-ino clusters. It is a research in the field of architectural design. The aim is to explore and reveal the architectural potentials of Le Corbusier&rsquo / s mass-producible Dom-ino frame (1914), not only by concentrating on a single Dom-ino unit but by focusing mainly to the clusters of Dom-ino, searching for the logic behind their multiplication, how they come together and to what extent the units allow variety. To achieve this, six alternative Dom-ino clusters are analyzed, and alternative Dom-ino units are explored, other than the well-known single one. Today, Dom-ino is argued as an architectural diagram, and this research reinforces this argument. The analysis of both the clusters and the units with different plan types eventually puts forward that the Dom-ino frame identifies and exploits all the potentials of mass-production with the standardization of the elements and modularity of the structural grid / and it is an adaptable, flexible and consequently a generic frame that produces infinitely alternative solutions. For this reason, this thesis argues that resolving the way Dom-ino units come together, and analyzing how and in what scale they allow variety in producing clusters will introduce the tools for proper mass-housing, and consequently could produce solutions for today&#039 / s mass-housing problems.

This thesis presents the computer implementation of a multi resolutional concurrent, design inference network, whose nodes are refined by PNDN (Petri Net Based Design Inference Network) modules. The extended design network is named as N-PNDN and consists of several embedded PNDN modules which models the information flow on a functional basis to facilitate the design automation at the conceptual design phase of an engineering design. Information flow in N-PNDN occurs between parent and child PNDN modules in a hierarchical structure and is provided by the token flow between the modules. In this study, computer implementation of the design network construction and token flow algorithms for the N-PNDN structure is restored and therefore the previous DNS (Design Network Simulator) is adapted for the multi layer design and decomposition of mechatronic products. The related algorithms are developed by using an object oriented, visual programming environment. The graphical user interface is also modified. The further developed DNS has been used for the application of the N-PNDN structure in the conceptual design of 5 mechatronic systems. In the guidance of this study, it has been understood that the further developed DNS is a powerful tool for designers coming from different engineering disciplines in order to interchange their ideas.