As redes ópticas estão passando por mudanças significativas, impulsionadas pelo crescimento exponencial do tráfego, principalmente advindo de serviços multimídia e armazenamento em nuvem. Esta demanda exigirá aumento da capacidade da taxa de transmissão para padrões como 400 Gb/s e 1 Tb/s. Nesse contexto, foi proposta uma arquitetura de rede com grade de frequências granular flexível chamada elastic optical network (EON). A EON divide o espectro de frequências em fatias (slots) de tamanho fixo e aloca grupos de slots contíguos estritamente de acordo com os requisitos de banda das demandas de conexão, implicando eficiência de uso do espectro. Com o aumento significativo da taxa de transmissão, acentuou-se a preocupação em manter a sobrevivência da rede, já que pouco tempo de queda no serviço pode acarretar uma imensa perda de dados. Neste trabalho, investigamos esquemas de proteção baseados em caminhos compartilhados (shared-path protection, SPP) e esquemas de restauração de tráfego. Avaliamos esquemas divulgados na literatura como o dynamic load balancing shared-path protection (DLBSPP) e esquemas de restauração como o traffic aware restoration (TAR) e bandwidth squeezed restoration (BSR). Avaliamos também uma heurística de alocação de slots chamada inverted dual stack (IDS). O DLBSPP utiliza balanceamento dinâmico de carga para computar os caminhos primários e de proteção compartilhados. O TAR executa a restauração dinâmica ordenando as conexões por granularidade de banda. O BSR utiliza a capacidade de contração de banda do EON para restaurar conexões por meio da política de melhor esforço ou de banda garantida, dependendo do acordo de níveis de serviço do cliente. O esquema IDS concentra o maior número possível de slots compartilhados em uma região do espectro. As medidas de desempenho dos algoritmos são avaliadas segundo as métricas: probabilidade de bloqueio, taxa de utilização do espectro, número médio de hops e taxa de restauração falha. As simulações computacionais mostram o bom desempenho da utilização do esquema IDS com DLBSPP. / Optical networks are undergoing significant changes driven by the exponentially growing traffic, especially coming from multimedia and cloud storage services. This demand will require increasing of the transmission rate capacity as high as 400 Gb/s and 1 Tb/s. Within this context, it was proposed the elastic optical network (EON), which is a network architecture with flexible granular frequency grid. EON divides the frequency spectrum into slices (slots) of fixed size and allocates groups of contiguous slots strictly according to the bandwidth requirement of the connection demands, providing high spectrum use efficiency. The significant increase in transmission rate put emphasis on the need to maintain the survival of the network, since the occurrence of faults in the network nodes or links can cause huge loss of data. In this work, we investigate protection schemes based on shared-path protection (SPP) and traffic restoration schemes. We evaluate schemes related in the literature, such as the dynamic load balancing shared-path protection (DLBSPP), and restoration schemes such as the traffic aware restoration (TAR) and the bandwidth squeezed restoration (BSR). The DLBP scheme uses dynamic load balancing to compute primary and shared protection paths. The TAR performs dynamic restoration ordering the connections based on band granularity. The BSR uses EON\'s band squeezing feature to restore connections by means of the best effort or guaranteed bandwidth strategy, depending on the customer\'s service level agreement. IDS scheme concentrates the maximum possible number of shared slots in a given region of the spectrum. Performance of the algorithms are evaluated according to metrics: blocking probability, spectrum utilization rate, average number of hops and failure restoration rate. Computer simulations show that the use of the IDS scheme improves the performance of the investigated algorithms.

Τα κατανεμημένα δίκτυα κλίμακας Ίντερνετ και κυρίως τα δίκτυα ομοτίμων εταίρων, γνωστά και ως peer-to-peer (p2p), που αποτελούν το πιο αντιπροσωπευτικό παράδειγμά τους, προσελκύουν τα τελευταία χρόνια μεγάλο ενδιαφέρον από τους ερευνητές και τις επιχειρήσεις λόγω των ιδιόμορφων χαρακτηριστικών τους, όπως ο πλήρης αποκεντρωτικός χαρακτήρας, η αυτονομία των κόμβων, η ικανότητα κλιμάκωσης, κ.λπ. Αρχικά σχεδιασμένα να υποστηρίζουν εφαρμογές διαμοιρασμού αρχείων με βασική υπηρεσία την επεξεργασία απλών ερωτημάτων, σύντομα εξελίχτηκαν σε ένα καινούργιο μοντέλο κατανεμημένων συστημάτων, με μεγάλες και αυξανόμενες δυνατότητες για διαδικτυακές εφαρμογές, υποστηρίζοντας πολύπλοκες εφαρμογές διαμοιρασμού δομημένων και σημασιολογικά προσδιορισμένων δεδομένων. Η προσέγγισή μας στην περιοχή αυτή γίνεται προς δύο βασικές κατευθύνσεις: (α) την επεξεργασία πολύπλοκων ερωτημάτων και (β) την εκτίμηση των ανομοιομορφιών των διαφόρων κατανομών που συναντάμε στα δίκτυα αυτά (π.χ. φορτίου, προσφοράς ή κατανάλωσης ενός πόρου, τιμών των δεδομένων των κόμβων, κ.λπ.), που εκτός των άλλων αποτελεί ένα σημαντικό εργαλείο στην υποστήριξη πολύπλοκων ερωτημάτων. Συγκεκριμένα, ασχολούμαστε και επιλύουμε τρία βασικά ανοικτά προβλήματα. Το πρώτο ανοικτό πρόβλημα είναι η επεξεργασία ερωτημάτων εύρους τιμών σε ομότιμα συστήματα κατανεμημένου πίνακα κατακερματισμού, με ταυτόχρονη εξασφάλιση της εξισορρόπησης του φορτίου των κόμβων και της ανοχής σε σφάλματα. Προτείνουμε μια αρχιτεκτονική επικάλυψης, που ονομάζουμε Saturn, που εφαρμόζεται πάνω από ένα δίκτυο κατανεμημένου πίνακα κατακερματισμού. Η αρχιτεκτονική Saturn χρησιμοποιεί: (α) μια πρωτότυπη συνάρτηση κατακερματισμού που τοποθετεί διαδοχικές τιμές δεδομένων σε γειτονικούς κόμβους, για την αποδοτική επεξεργασία των ερωτημάτων εύρους τιμών και (β) την αντιγραφή, για την εξασφάλιση της εξισορρόπησης του φορτίου προσπελάσεων (κάθετη, καθοδηγούμενη από το φορτίο αντιγραφή) και της ανοχής σε σφάλματα (οριζόντια αντιγραφή). Μέσα από μια εκτεταμένη πειραματική αξιολόγηση του Saturn και σύγκριση με δύο βασικά δίκτυα κατανεμημένου πίνακα κατακερματισμού (Chord και OP-Chord) πιστοποιούμε την ανωτερότητα του Saturn να αντιμετωπίζει και τα τρία ζητήματα που θέσαμε, αλλά και την ικανότητά του να συντονίζει το βαθμό αντιγραφής ώστε να ανταλλάζει ανάμεσα στο κόστος αντιγραφής και στο βαθμό εξισορρόπησης του φορτίου. Το δεύτερο ανοικτό πρόβλημα που αντιμετωπίζουμε αφορά την έλλειψη κατάλληλων μετρικών που να εκφράζουν τις ανομοιομορφίες των διαφόρων κατανομών (όπως, για παράδειγμα, το βαθμό δικαιοσύνης μιας κατανομής φορτίου) σε κατανεμημένα δίκτυα κλίμακας Ίντερνετ και την μη αποτελεσματική ή δυναμική εκμετάλλευση μετρικών ανομοιομορφίας σε συνδυασμό με αλγορίθμους διόρθωσης (όπως ο αλγόριθμος εξισορρόπησης φορτίου). Το πρόβλημα είναι σημαντικό γιατί η εκτίμηση των κατανομών συντελεί στην ικανότητα κλιμάκωσης και στην επίδοση αυτών των δικτύων. Αρχικά, προτείνουμε τρεις μετρικές ανομοιομορφίας (το συντελεστή του Gini, τον δείκτη δικαιοσύνης και το συντελεστή διασποράς) μετά από μια αναλυτική αξιολόγηση μεταξύ γνωστών μετρικών εκτίμησης ανομοιομορφίας και στη συνέχεια, αναπτύσσουμε τεχνικές δειγματοληψίας (τρεις γνωστές τεχνικές και τρεις προτεινόμενες) για τη δυναμική εκτίμηση αυτών των μετρικών. Με εκτεταμένα πειράματα αξιολογούμε συγκριτικά τους προτεινόμενους αλγορίθμους εκτίμησης και τις τρεις μετρικές και επιδεικνύουμε πώς αυτές οι μετρικές και ειδικά, ο συντελεστής του Gini, μπορούν να χρησιμοποιηθούν εύκολα και δυναμικά από υψηλότερου επιπέδου αλγορίθμους, οι οποίοι μπορούν τώρα να ξέρουν πότε να επέμβουν για να διορθώσουν τις άδικες κατανομές. Το τρίτο και τελευταίο ανοικτό πρόβλημα αφορά την εκτίμηση του μεγέθους αυτοσύνδεσης μιας σχέσης όπου οι πλειάδες της είναι κατανεμημένες σε κόμβους δεδομένων που αποτελούν ένα ομότιμο δίκτυο επικάλυψης. Το μέγεθος αυτοσύνδεσης έχει χρησιμοποιηθεί εκτεταμένα σε συγκεντρωτικές βάσεις δεδομένων για τη βελτιστοποίηση ερωτημάτων και υποστηρίζουμε ότι μπορεί να χρησιμοποιηθεί και σε ένα πλήθος άλλων εφαρμογών, ειδικά στα ομότιμα δίκτυα (π.χ. συσταδοποίηση του Ιστού, αναζήτηση στον Ιστό, κ.λπ.). Η συνεισφορά μας περιλαμβάνει, αρχικά, τις προσαρμογές πέντε γνωστών συγκεντρωτικών τεχνικών εκτίμησης του μεγέθους αυτοσύνδεσης (συγκεκριμένα, σειριακή, ετεροδειγματοληπτική, προσαρμοστική και διεστιακή δειγματοληψία και δειγματοληψία με μέτρηση δείγματος) στο περιβάλλον ομοτίμων εταίρων και η ανάπτυξη μια πρωτότυπης τεχνικής εκτίμησης του μεγέθους αυτοσύνδεσης, βασισμένη στο συντελεστή του Gini. Με μαθηματική ανάλυση δείχνουμε ότι οι εκτιμήσεις του συντελεστή του Gini μπορούν να οδηγήσουν σε εκτιμήσεις των υποκείμενων κατανομών δεδομένων, όταν αυτά ακολουθούν το νόμο της δύναμης ή το νόμο του Zipf και αυτές, με τη σειρά τους, σε εκτιμήσεις του μεγέθους αυτοσύνδεσης των σχέσεων των δεδομένων. Μετά από αναλυτική πειραματική μελέτη και σύγκριση όλων των παραπάνω τεχνικών αποδεικνύουμε ότι η καινούργια τεχνική που προτείνουμε είναι πολύ αποτελεσματική ως προς την ακρίβεια, την πιστότητα και την απόδοση έναντι των άλλων πέντε μεθόδων. / The distributed, Internet-scale networks, and mainly, the peer-to-peer networks (p2p), that constitute their most representative example, recently attract a great interest from the researchers and the industry, due to their outstanding properties, such as full decentralization, autonomy of nodes, scalability, etc. Initially designed to support file sharing applications with simple lookup operations, they soon developed in a new model of distributed systems, with many and increasing possibilities for Internet applications, supporting complex applications of structured and semantically rich data. Our research to the area has two basic points of view: (a) complex query processing and (b) estimation of skewness in various distributions existing in these networks (e.g. load distribution, distribution of offer, or consumption of resources, data value distributions, etc), which, among others, it is an important tool to complex query processing support. Specifically, we deal with and solve three basic open problems. The first open problem is range query processing in p2p systems based on distributed hash tables (DHT), with simultaneous guarantees of access load balancing and fault tolerance. We propose an overlay DHT architecture, coined Saturn. Saturn uses a novel order-preserving hash function that places consecutive data values in successive nodes to provide efficient range query processing, and replication to guarantee access load balancing (vertical, load-driven replication) and fault tolerance (horizontal replication). With extensive experimentation, we evaluate and compare Saturn with two basic DHT networks (Chord and OP - Chord), and certify its superiority to cope with the three above requirements, but also its ability to tune the degree of replication to trade off replication costs for access load balancing. The second open problem that we face concerns the lack of appropriate metrics to express the degree of skewness of various distributions (for example, the fairness degree of load balancing) in p2p networks, and the inefficient and offline-only exploitation of metrics of skewness, which does not enable any cooperation with corrective algorithms (for example, load balancing algorithms). The problem is important because estimation of distribution fairness contributes to system scalability and efficiency. First, after a comprehensive study and evaluation of popular metrics of skewness, we propose three of them (the coefficient of Gini, the fairness index, and the coefficient of variation), and, then, we develop sampling techniques (three already known techniques, and three novel ones) to dynamically estimate these metrics. With extensive experimentation, which comparatively evaluates both the various proposed estimation algorithms and the three metrics we propose, we show how these three metrics, and especially, the coefficient of Gini, can be easily utilized online by higher-level algorithms, which can now know when to best intervene to correct unfairness. The third and last open problem concerns self-join size estimation of a relation whose tuples are distributed over data nodes which comprise an overlay network. Self-join size has been extensively used in centralized databases for query optimization purposes, and we support that it can also be used in various other applications, specifically in p2p networks (e.g. web clustering, web searching, etc). Our contribution first includes the adaptations of five well-known self-join size estimation, centralized techniques (specifically, sequential sampling, cross-sampling, adaptive and bifocal sampling, and sample-count) to the p2p environment and a novel estimation technique which is based on the Gini coefficient. With mathematical analysis we show that, the estimates of the Gini coefficient can lead to estimates of the degree of skewness of the underlying data distribution, when these follow the power, or Zipf’s law, and these estimates can lead to self-join size estimates of those data relations. With extensive experimental study and comparison of all above techniques, we prove that the proposed technique is very efficient in terms of accuracy, precision, and cost of estimation against the other five methods.

Submitted by Marlene Santos (marlene.bc.ufg@gmail.com) on 2017-09-29T20:50:34Z No. of bitstreams: 2 Dissertação - Vínicius Nunes Medeiros - 2017.pdf: 3262332 bytes, checksum: 6481a1cb0287306368b482913e18928d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2017-10-02T12:58:17Z (GMT) No. of bitstreams: 2 Dissertação - Vínicius Nunes Medeiros - 2017.pdf: 3262332 bytes, checksum: 6481a1cb0287306368b482913e18928d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2017-10-02T12:58:17Z (GMT). No. of bitstreams: 2 Dissertação - Vínicius Nunes Medeiros - 2017.pdf: 3262332 bytes, checksum: 6481a1cb0287306368b482913e18928d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2017-09-01 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Owing to the huge range of application scenarios for wireless networks, new prospects for the current and future applications of the Internet, making new communication paradigms such as the Internet of Things (IoT) and 5G networks. Two important networks are included in this scenery: Wireless Sensor Networks (WSN) provide key devices for developing the IoT communication paradigm, such as the sensors used for collecting different kind of information, and the Wireless Mesh Network (WMN) that enable interoperability between heterogeneous network, covering large areas with at am low cost.Eeach one these networks have specific characteristics, for example, the WSN has serious restrictions (e.g. power consumption, low speed wireless links, interference) while the WMN offers high data rates through multiple radios and channels. In this context, it is necessary to find solutions that can ensure more efficient communication system based on the optimized utilization of the network resources. So it was created a multi-objective routing algorithm, called Routing-Aware of path Length, Link quality, and traffic Load (RALL). This seeks to strike a balance between three objectives: to minimize the network bottlenecks, to reduce path length, and to avoid links with low quality, for resolution the routing problem for WSN, Then the RALL approach was adapted to match the specialty of the WMN, entitled Routing Aware of path Length, Link quality, and traffic Load Weighted (RALLW). The results indicate that combining multiple objectives when performing routing can benefit networks performance. The RALL and RALLW results in a good performance with regard to the delivery rate, throughput, delay, and power consumption when compared to other approaches, however the RALLW approach did not significant differences when compared to the RALL. / A utilização das redes sem fio em diversas aplicações está criando uma nova perspectiva de utilidade para a Internet atual fazendo com que novos paradigmas de comunicação sejam desenvolvidos, como a Internet das coisas (IoT) e redes 5G. Dois tipos de rede sem fio se destacam nesses novos cenários, a Rede de Sensores Sem Fio (RSSF) que fornece dispositivos fundamentais para a implementação dos ambientes inteligentes e da IoT, e a Rede em Malha Sem Fio (RMSF) que permite a interoperabilidade entre os tipos de redes, cobrindo grandes áreas e com um custo relativamente baixo. Essas redes possuem peculiaridade específicas, por exemplo a RSSF possui fortes restrições de consumo energético e capacidade de transmissão, já a RMSF é uma rede com maior capacidade de transmissão e maior alcance. Nesse contexto, é necessário desenvolver soluções que tornam a rede eficiente através da otimização dos seus recursos. Inicialmente foi desenvolvida a abordagem Routing Aware of path Length, Link quality, and traffic Load (RALL) para realizar o roteamento multiobjetivo em RSSF, em seguida, realizamos uma adaptação dessa abordagem para refletir as especificidades dos cenários de uma RMSF, essa adaptação foi intitulada de Routing Aware of path Length, Link quality, and traffic Load Weigthed (RALLW). Em ambas abordagens o problema de roteamento multiobjetivo é composto por três objetivos (quantidade de saltos, quantidade de enlaces com baixa qualidade e balanceamento da rede), comuns a RSSF e a RMSF, apesar de existirem diferenças significativas entre elas ao modelar os objetivos. Os resultados indicam que a combinação de múltiplos objetivos ao realizar o roteamento pode beneficiar o desempenho da rede. As abordagens RALL e RALLW resultaram em bons resultados quando comparados com outras abordagens, em relação à taxa de entrega, vazão, atraso e consumo de energia, entretanto, a abordagem RALLW não mostrou diferenças significativas quando comparada a abordagem RALL.

As redes ópticas estão passando por mudanças significativas, impulsionadas pelo crescimento exponencial do tráfego, principalmente advindo de serviços multimídia e armazenamento em nuvem. Esta demanda exigirá aumento da capacidade da taxa de transmissão para padrões como 400 Gb/s e 1 Tb/s. Nesse contexto, foi proposta uma arquitetura de rede com grade de frequências granular flexível chamada elastic optical network (EON). A EON divide o espectro de frequências em fatias (slots) de tamanho fixo e aloca grupos de slots contíguos estritamente de acordo com os requisitos de banda das demandas de conexão, implicando eficiência de uso do espectro. Com o aumento significativo da taxa de transmissão, acentuou-se a preocupação em manter a sobrevivência da rede, já que pouco tempo de queda no serviço pode acarretar uma imensa perda de dados. Neste trabalho, investigamos esquemas de proteção baseados em caminhos compartilhados (shared-path protection, SPP) e esquemas de restauração de tráfego. Avaliamos esquemas divulgados na literatura como o dynamic load balancing shared-path protection (DLBSPP) e esquemas de restauração como o traffic aware restoration (TAR) e bandwidth squeezed restoration (BSR). Avaliamos também uma heurística de alocação de slots chamada inverted dual stack (IDS). O DLBSPP utiliza balanceamento dinâmico de carga para computar os caminhos primários e de proteção compartilhados. O TAR executa a restauração dinâmica ordenando as conexões por granularidade de banda. O BSR utiliza a capacidade de contração de banda do EON para restaurar conexões por meio da política de melhor esforço ou de banda garantida, dependendo do acordo de níveis de serviço do cliente. O esquema IDS concentra o maior número possível de slots compartilhados em uma região do espectro. As medidas de desempenho dos algoritmos são avaliadas segundo as métricas: probabilidade de bloqueio, taxa de utilização do espectro, número médio de hops e taxa de restauração falha. As simulações computacionais mostram o bom desempenho da utilização do esquema IDS com DLBSPP. / Optical networks are undergoing significant changes driven by the exponentially growing traffic, especially coming from multimedia and cloud storage services. This demand will require increasing of the transmission rate capacity as high as 400 Gb/s and 1 Tb/s. Within this context, it was proposed the elastic optical network (EON), which is a network architecture with flexible granular frequency grid. EON divides the frequency spectrum into slices (slots) of fixed size and allocates groups of contiguous slots strictly according to the bandwidth requirement of the connection demands, providing high spectrum use efficiency. The significant increase in transmission rate put emphasis on the need to maintain the survival of the network, since the occurrence of faults in the network nodes or links can cause huge loss of data. In this work, we investigate protection schemes based on shared-path protection (SPP) and traffic restoration schemes. We evaluate schemes related in the literature, such as the dynamic load balancing shared-path protection (DLBSPP), and restoration schemes such as the traffic aware restoration (TAR) and the bandwidth squeezed restoration (BSR). The DLBP scheme uses dynamic load balancing to compute primary and shared protection paths. The TAR performs dynamic restoration ordering the connections based on band granularity. The BSR uses EON\'s band squeezing feature to restore connections by means of the best effort or guaranteed bandwidth strategy, depending on the customer\'s service level agreement. IDS scheme concentrates the maximum possible number of shared slots in a given region of the spectrum. Performance of the algorithms are evaluated according to metrics: blocking probability, spectrum utilization rate, average number of hops and failure restoration rate. Computer simulations show that the use of the IDS scheme improves the performance of the investigated algorithms.

Dans cette thèse, nous nous intéressons à la résolution parallèle de grands systèmes linéaires creux. Nous nous focalisons plus particulièrement sur les solveurs linéaires creux hybrides directs itératifs tels que HIPS, MaPHyS, PDSLIN ou ShyLU, qui sont basés sur une décomposition de domaine et une approche « complément de Schur ». Bien que ces solveurs soient moins coûteux en temps et en mémoire que leurs homologues directs, ils ne sont néanmoins pas exempts de surcoûts. Dans une première partie, nous présentons les différentes méthodes de réduction de la consommation mémoire déjà existantes et en proposons une nouvelle qui n’impacte pas la robustesse numérique du précondionneur construit. Cette technique se base sur une atténuation du pic mémoire par un ordonnancement spécifique des tâches de calcul, d’allocation et de désallocation des blocs, notamment ceux se trouvant dans les parties « couplage » des domaines.Dans une seconde partie, nous nous intéressons à la question de l’équilibrage de la charge que pose la décomposition de domaine pour le calcul parallèle. Ce problème revient à partitionner le graphe d’adjacence de la matrice en autant de parties que de domaines désirés. Nous mettons en évidence le fait que pour avoir un équilibrage correct des temps de calcul lors des phases les plus coûteuses d’un solveur hybride tel que MaPHyS, il faut à la fois équilibrer les domaines en termes de nombre de noeuds et de taille d’interface locale. Jusqu’à aujourd’hui, les partitionneurs de graphes tels que Scotch et MeTiS ne s’intéressaient toutefois qu’au premier critère (la taille des domaines) dans le contexte de la renumérotation des matrices creuses. Nous proposons plusieurs variantes des algorithmes existants afin de prendre également en compte l’équilibrage des interfaces locales. Toutes nos modifications sont implémentées dans le partitionneur Scotch, et nous présentons des résultats sur de grands cas de tests industriels. / In this thesis, we focus on the parallel solving of large sparse linear systems. Our main interestis on direct-iterative hybrid solvers such as HIPS, MaPHyS, PDSLIN or ShyLU, whichrely on domain decomposition and Schur complement approaches. Althrough these solvers arenot as time and space consuming as direct methods, they still suffer from serious overheads. Ina first part, we thus present the existing techniques for reducing the memory consumption, andwe present a new method which does not impact the numerical robustness of the preconditioner.This technique reduces the memory peak by doing a special scheduling of computation, allocation,and freeing tasks in particular in the Schur coupling blocks of the matrix. In a second part,we focus on the load balancing of the domain decomposition in a parallel context. This problemconsists in partitioning the adjacency graph of the matrix in as many domains as desired. Wepoint out that a good load balancing for the most expensive steps of an hybrid solver such asMaPHyS relies on the balancing of both interior nodes and interface nodes of the domains.Through, until now, graph partitioners such as MeTiS or Scotch used to optimize only thefirst criteria (i.e., the balancing of interior nodes) in the context of sparse matrix ordering. Wepropose different variations of the existing algorithms to improve the balancing of interface nodesand interior nodes simultaneously. All our changes are implemented in the Scotch partitioner.We present our results on large collection of matrices coming from real industrial cases.

Energy consumption is a major concern in Wireless Sensor Networks (WSNs) resulting in a strong demand for energy-aware communication technologies. In this context, several unequal cluster-based routing protocols have been proposed. However, few of them adopt energetic analysis models for the calculation of the optimal cluster radius and several protocols can not realize an optimal workload balance between sensor nodes. In this scope, the aim of the dissertation is to develop a cluster-based routing protocol for improving energy efficiency in WSN. We propose a Fuzzy-based Energy-Aware Unequal Clustering algorithm (FEAUC) with circular partitioning to balance the energy consumption between sensor nodes and solve the hotspot problem created by a multi-hop communication. The developed FEAUC involves mainly four phases: An off-line phase, a cluster formation phase, a cooperation phase and data collection phase. During the off-line phase, an energy analysis is performed to calculate the radius of each ring and the optimal cluster radius per ring. The cluster formation phase is based on a fuzzy logic approach for the cluster head (CH) selection. The cooperation phase aims to define an intermediate node as a router between different CHs. While, in the data collection phase, transmitting data packet from sensor nodes to their appropriate CHs is defined as an intra-cluster communication, and transmitting data from one CH to another until reaching the base station, is defined as an inter-cluster communication. The feasibility of the developed FEAUC is demonstrated by elaborating comparison with selected referred unequal clustering algorithms considering different parameters, mainly, the energy consumption, battery lifetime, time to first node shuts down (FND), time of half of nodes off-line (HND) and time to last node dies (LND). Although, the developed FEAUC is intended to enhance the network lifetime by distributing the large load of CH tasks equally among the normal nodes, running the clustering process in each round is an additional burden, which can significantly drain the remaining energy. For this reason, the FEAUC based protocol has been further developed to become a fault tolerant algorithm (FEAUC-FT). It supports the fault tolerance by using backup CHs to avoid the re-clustering process in certain rounds or by building further routing paths in case of a link failure between different CHs. The validation of the developed FEAUC in real scenarios has been performed. Some sensor nodes, powered with batteries, are deployed in a circular area forming clusters. Performance evaluations are carried out by realistic scenarios and tested for a real deployment using the low-power wireless sensor node panStamp. To complete previous works, as a step of proof of concept, a smart irrigation system is designed, called Air-IoT. Furthermore, a real-time IoT-based sensor node architecture to control the quantity of water in some deployed nodes is introduced. To this end, a cloud-connected wireless network to monitor the soil moisture and temperature is well-designed. Generally, this step is essential to validate and evaluate the proposed unequal cluster-based routing algorithm in a real demonstrator. The proposed prototype guarantees both real-time monitoring and reliable and cost-effective transmission between each node and the base station.:1 Introduction 2 Theoretical background 3 State of the art of unequal cluster-based routing protocols 4 FEAUC: Fuzzy-based Energy-Aware Unequal Clustering 5 Experimental validation of the developed unequal clustering protocol 6 Real application to specific uses cases 7 Conclusions and future research directions / Der Energieverbrauch ist ein Hauptanliegen in drahtlosen Sensornetzwerken (WSNs), was zu einer starken Nachfrage nach energiebewussten Kommunikationstechnologien führt. In diesem Zusammenhang wurden mehrere ungleiche clusterbasierte Routing-Protokolle vorgeschlagen. Allerdings verwenden nur die wenigsten energetische Analysemodelle für die Berechnung des optimalen Cluster-Radius, und mehrere Protokolle können keine optimale Auslastungsbalance zwischen Sensorknoten realisieren. In diesem Zusammenhang ist es das Ziel der Dissertation, ein clusterbasiertes Routing-Protokoll zur Verbesserung der Energieeffizienz im WSN zu entwickeln. Wir schlagen einen Fuzzy-basierten Energy-Aware Unequal Clustering-Algorithmus (FEAUC) mit zirkulärer Partitionierung vor, um den Energieverbrauch zwischen Sensorknoten auszugleichen und das durch eine Multi-Hop-Kommunikation entstehende Hotspot-Problem zu lösen. Der entwickelte FEAUC umfasst hauptsächlich vier Phasen: Eine Offline-Phase, eine Clusterbildungsphase, eine Kooperationsphase und eine Phase der Datensammlung. Während der Offline-Phase wird eine Energieanalyse durchgeführt, um den Radius jedes Ringes und den optimalen Cluster- Radius pro Ring zu berechnen. Die Clusterbildungsphase basiert auf einem Fuzzy-Logik-Ansatz für die Clusterkopf (CH)-Auswahl. Die Kooperationsphase zielt darauf ab, einen Zwischenknoten als einen Router zwischen verschiedenen CHs zu definieren. In der Datensammelphase wird die Übertragung von Datenpaketen von Sensorknoten zu ihren entsprechenden CHs als eine Intra-Cluster-Kommunikation definiert, während die Übertragung von Daten von einem CH zu einem anderen CH bis zum Erreichen der Basisstation als eine Inter-Cluster-Kommunikation definiert wird. Die Machbarkeit des entwickelten FEAUC wird durch die Ausarbeitung eines Vergleichs mit ausgewählten referenzierten ungleichen Clustering-Algorithmen unter Berücksichtigung verschiedener Parameter demonstriert, hauptsächlich des Energieverbrauchs, der Batterielebensdauer, der Zeit bis zum Abschalten des ersten Knotens (FND), der Zeit, in der die Hälfte der Knoten offline ist (HND) und der Zeit bis zum letzten Knoten stirbt (LND). Obwohl mit dem entwickelten FEAUC die Lebensdauer des Netzwerks erhöht warden soll, indem die große Last der CH-Aufgaben gleichmäßig auf die übrigen Knoten verteilt wird, stellt die Durchführung des Clustering-Prozesses in jeder Runde eine zusätzliche Belastung dar, die die verbleibende Energie erheblich entziehen kann. Aus diesem Grund wurde das auf FEAUC basierende Protokoll zu einem fehlerto-leranten Algorithmus (FEAUC-FT) weiterentwickelt. Er unterstützt die Fehlerto-leranz durch die Verwendung von Backup-CHs zur Vermeidung des Re-Clustering-Prozesses in bestimmten Runden oder durch den Aufbau weiterer Routing-Pfade im Falle eines Verbindungsausfalls zwischen verschiedenen CHs. Die Validierung des entwickelten FEAUC in realen Szenarien ist durchgeführt worden. Einige Sensorknoten, die mit Batterien betrieben werden, sind in einem kreisförmigen Bereich angeordnet und bilden Cluster. Leistungsbewertungen warden anhand realistischer Szenarien durchgeführt und für einen realen Einsatz unter Verwendung des drahtlosen Low-Power-Sensorknoten panStamp getestet. Zur Vervollständigung früherer Arbeiten wird als Schritt des Proof-of-Concept ein intelligentes Bewässerungssystem mit der Bezeichnung Air-IoT entworfen. Darüber hinaus wird eine IoT-basierte Echtzeit-Sensorknotenarchitektur zur Kontrolle derWassermenge in einigen eingesetzten Knoten eingeführt. Zu diesem Zweck wird ein mit der Cloud verbundenes drahtloses Netzwerk zur Überwachung der Bodenfeuchtigkeit und -temperatur gut konzipiert. Im Allgemeinen ist dieser Schritt unerlässlich, um den vorgeschlagenen ungleichen clusterbasierten Routing-Algorithmus in einem realen Demonstrator zu validieren und zu bewerten.Der vorgeschlagene Prototyp garantiert sowohl Echtzeit-Überwachung als auch zuverlässige und kostengünstige Übertragung zwischen jedem Knoten und der Basisstation.:1 Introduction 2 Theoretical background 3 State of the art of unequal cluster-based routing protocols 4 FEAUC: Fuzzy-based Energy-Aware Unequal Clustering 5 Experimental validation of the developed unequal clustering protocol 6 Real application to specific uses cases 7 Conclusions and future research directions

Over the last years, the Linked Open Data (LOD) has evolved from a mere 12 to more than 10,000 knowledge bases. These knowledge bases come from diverse domains including (but not limited to) publications, life sciences, social networking, government, media, linguistics. Moreover, the LOD cloud also contains a large number of crossdomain knowledge bases such as DBpedia and Yago2. These knowledge bases are commonly managed in a decentralized fashion and contain partly verlapping information. This architectural choice has led to knowledge pertaining to the same domain being published by independent entities in the LOD cloud. For example, information on drugs can be found in Diseasome as well as DBpedia and Drugbank. Furthermore, certain knowledge bases such as DBLP have been published by several bodies, which in turn has lead to duplicated content in the LOD . In addition, large amounts of geo-spatial information have been made available with the growth of heterogeneous Web of Data. The concurrent publication of knowledge bases containing related information promises to become a phenomenon of increasing importance with the growth of the number of independent data providers. Enabling the joint use of the knowledge bases published by these providers for tasks such as federated queries, cross-ontology question answering and data integration is most commonly tackled by creating links between the resources described within these knowledge bases. Within this thesis, we spur the transition from isolated knowledge bases to enriched Linked Data sets where information can be easily integrated and processed. To achieve this goal, we provide concepts, approaches and use cases that facilitate the integration and enrichment of information with other data types that are already present on the Linked Data Web with a focus on geo-spatial data. The first challenge that motivates our work is the lack of measures that use the geographic data for linking geo-spatial knowledge bases. This is partly due to the geo-spatial resources being described by the means of vector geometry. In particular, discrepancies in granularity and error measurements across knowledge bases render the selection of appropriate distance measures for geo-spatial resources difficult. We address this challenge by evaluating existing literature for point set measures that can be used to measure the similarity of vector geometries. Then, we present and evaluate the ten measures that we derived from the literature on samples of three real knowledge bases. The second challenge we address in this thesis is the lack of automatic Link Discovery (LD) approaches capable of dealing with geospatial knowledge bases with missing and erroneous data. To this end, we present Colibri, an unsupervised approach that allows discovering links between knowledge bases while improving the quality of the instance data in these knowledge bases. A Colibri iteration begins by generating links between knowledge bases. Then, the approach makes use of these links to detect resources with probably erroneous or missing information. This erroneous or missing information detected by the approach is finally corrected or added. The third challenge we address is the lack of scalable LD approaches for tackling big geo-spatial knowledge bases. Thus, we present Deterministic Particle-Swarm Optimization (DPSO), a novel load balancing technique for LD on parallel hardware based on particle-swarm optimization. We combine this approach with the Orchid algorithm for geo-spatial linking and evaluate it on real and artificial data sets. The lack of approaches for automatic updating of links of an evolving knowledge base is our fourth challenge. This challenge is addressed in this thesis by the Wombat algorithm. Wombat is a novel approach for the discovery of links between knowledge bases that relies exclusively on positive examples. Wombat is based on generalisation via an upward refinement operator to traverse the space of Link Specifications (LS). We study the theoretical characteristics of Wombat and evaluate it on different benchmark data sets. The last challenge addressed herein is the lack of automatic approaches for geo-spatial knowledge base enrichment. Thus, we propose Deer, a supervised learning approach based on a refinement operator for enriching Resource Description Framework (RDF) data sets. We show how we can use exemplary descriptions of enriched resources to generate accurate enrichment pipelines. We evaluate our approach against manually defined enrichment pipelines and show that our approach can learn accurate pipelines even when provided with a small number of training examples. Each of the proposed approaches is implemented and evaluated against state-of-the-art approaches on real and/or artificial data sets. Moreover, all approaches are peer-reviewed and published in a conference or a journal paper. Throughout this thesis, we detail the ideas, implementation and the evaluation of each of the approaches. Moreover, we discuss each approach and present lessons learned. Finally, we conclude this thesis by presenting a set of possible future extensions and use cases for each of the proposed approaches.

Controlling the balance of production and consumption of electricity will become increasingly challenging as the transport sector gradually converts to electric vehicles along with a growing share of wind power in the Swedish electric power system. This puts greater demand on resources that maintain the balance to ensure stable grid operation. The balancing act is called frequency regulation which historically has been performed almost entirely by hydropower. As the power production becomes more intermittent with renewable energy sources, frequency regulation will need to be performed in higher volumes on the demand side by having a more flexible consumption. In this report, the electrification of 17 buses Svealandstrafiken bus depot in Västerås has been studied. The aim has been to assess different charging strategies to efficiently utilize the available time and power but also to investigate if Svealandstrafiken can participate in frequency regulation. A smart charging model was created that demonstrated how smart charging can be implemented to optimize the charging in four different cases. The simulated cases were: charging with load balancing, reduced charging power, frequency regulation, and electrifying more buses. The results show that the power capacity limit will be exceeded if the buses are being charged directly as they arrive at the depot and without scheduling the charging session. By implementing smart charging, Svealandstrafiken can fully charge the 17 buses within the power capacity limit of the depot with 82 minutes to spare. By utilizing this 82-minute margin in the four different charging strategies, it was found that Svealandstrafiken can save 88 200SEK per year by load balancing, save 30 000 SEK per year by reducing the charging power by 10 %, earn 111 900 SEK per year by frequency regulation or electrify five more buses. Reducing the charging power may also increase the lifetime of the batteries but quantifying this needs further studies. Conclusively, there is economic potential for Svealandstrafiken for implementing smart charging.

Les applications multimédias, dans les réseaux de capteurs sans fil, véhiculent des données volumineuses, qui nécessitent un taux de transmission élevé et un traitement intensif et par conséquent une consommation d'énergie importante. Transmettre efficacement ces flux hétérogènes, tout en assurant leur fiabilité et garantir les exigences de QoS, avec les ressources limitées disponibles, en particulier dans les contextes critiques, demeure un verrou scientifique ouvert. C’est pourquoi, dans le cadre de cette thèse, nous nous sommes intéressés aux aspects liés : au routage, à la tolérance aux pannes et à la gestion des interférences dans les RCMSF.Compte tenu, du très faible nombre de protocoles de routage, qui ont approché la phase expérimentale et encore moins ceux qui l’ont abordé pendant la transmission des données en temps réel, nous avons développé dans ce contexte, un protocole routage géographique baptisé GNMFT (Geographic Non-interfering Multipath Fault-tolerant),En premier lieu, nous avons amélioré le mode de sélection des nœuds, utilisé par le transfert glouton, pour faire face au problème du Minimum-local, où une fonction objective multicritères (distance, énergie et puissance de réception) relative au choix du prochain saut a été définie. Puis, nous avons introduit une phase d’optimisation des chemins construits afin d’éliminer les boucles et une approche adaptative gérant les transmissions simultanées des différents trafics.Par la suite, nous nous sommes orientés vers la tolérance aux pannes, pour assurer la fiabilité des données transmises ainsi que la connectivité du réseau. A cet effet, nous avons modélisé le nombre des paquets perdus durant la livraison des paquets et proposé deux mécanismes : un curatif pour réparer les défaillances soudaines et un préventif afin d’anticiper l’épuisement des batteries. Les deux mécanismes sont combinés avec une stratégie de basculement dynamique lors de la construction des chemins alternatives.En dernier lieu, nous avons présenté un modèle d’interférence et un troisième mécanisme qui limite les interactions entre les liens adjacents. De plus, nous avons défini également un modèle de perte de chemins dans un environnement multi-paires (source,sink) et nous avons calculé le coefficient de probabilité d’erreurs basée sur une fonction de distance qui sépare les flux de données. Une nouvelle métrique a été ajouté dans la fonction objective, relative à la somme des bruits des nœuds actifs qui interférent sur les nœuds du ForwardingSet du nœud courant.Les résultats obtenus montrent l’efficacité des approches proposées qui ont été étudiées et validées à la fois par simulation et sur un banc d’essai expérimental. / Multimedia applications in WSNs convey large data (image, audio and video) that requires high transmission rate and intensive treatment and therefore high energy consumption. Effectively transmit these heterogeneous flows, while ensuring their reliability and guaranteeing QoS requirements, with the limited resources available, especially in critical contexts, remains an open scientific problem. That is why, in this thesis, we are interested in aspects related to : routing, fault tolerance and interference management in WMSNs.Given the very low number of routing protocols, that have approached the experimental phase and still less those who approached it during data transmission in real time, we developed in this context, a geographic routing protocol baptised GNMF (Geographic Non-interfering Multipath Fault-tolerant),First, we improved node's selection mode used by the greedy-forwarding, to deal with local minimum problem, where a multi-criteria objective function (distance, energy and reception power) related to next-hop choice has been defined. Then, we introduced an optimization phase of built paths to eliminate loops and an adaptive approach to manage simultaneous traffic transmissions.After that, we oriented towards fault tolerance, to ensure transmitted data reliability and network connectivity. To this end, we modeled the number of lost packets during package delivery and proposed two mechanisms. The curative is used when sudden failures occurs and the preventive to anticipate batteries depletion. Both are combined with a dynamic failover strategy during alternative paths construction.Finally, we presented an interference model and a third mechanism that limits interactions between adjacent links. In addition, we also defined a path loss model in a multipairs environment (source, sink) and computed the error probability coefficient based on a distance function that separates the data flows. A new metric has been added in the objective function, related to noise sum of the active nodes that interferes on forwarding set nodes of the current node.Obtained results show the effectiveness of the proposed approaches that have been studied and validated both by simulation and on an experimental testbed.

Les attaques massives par déni de service représentent une menace pour les services Internet. Ils impactent aussi les fournisseurs de service réseau et menace même la stabilité de l’Internet. Il y a donc un besoin pressant de contrôler les dommages causés par ces attaques. De nombreuses recherches ont été menées, mais aucune n’a été capable de combiner le besoin d’atténuation de l’attaque, avec l’obligation de continuité de service et les contraintes réseau. Les contre mesures proposées portent sur l’authentification des clients légitimes, le filtrage du trafic malicieux, une utilisation efficace des interconnections entre les équipements réseaux, ou l’absorption de l’attaque par les ressources disponibles. Dans cette thèse, nous proposons un mécanisme de contrôle de dommages. Basé sur une nouvelle signature d’attaque et les fonctions réseaux du standard Multiprotocol Label Switching (MPLS), nous isolons le trafic malicieux du trafic légitime et appliquons des contraintes sur la transmission du trafic malicieux. Le but est de rejeter suffisamment de trafic d’attaque pour maintenir la stabilité du réseau tout en préservant le trafic légitime. La solution prend en compte des informations sur l’attaque, mais aussi les ressources réseaux. Considérant que les opérateurs réseaux n’ont pas une même visibilité sur leur réseau, nous étudions l’impact de contraintes opérationnelles sur l’efficacité d’une contre mesure régulièrement recommandée, le filtrage par liste noire. Les critères d’évaluation sont le niveau d’information sur l’attaque ainsi que sur le trafic réseau. Nous formulons des scénarios auxquels chaque opérateur peut s’identifier. Nous démontrons que la l’algorithme de génération des listes noires doit être choisi avec précaution afin de maximiser l’efficacité du filtrage / Massive Denial of Service attacks represent a genuine threat for Internet service, but also significantly impact network service providers and even threat the Internet stability. There is a pressing need to control damages caused by such attacks. Numerous works have been carried out, but were unable to combine the need for mitigation, the obligation to provide continuity of service and network constraints. Proposed countermeasures focus on authenticating legitimate traffic, filtering malicious traffic, making better use of interconnection between network equipment or absorbing attack with the help of available resources. In this thesis, we propose a damage control mechanism against volumetric Denial of Services. Based on a novel attack signature and with the help of Multiprotocol Label Switching (MPLS) network functions, we isolate malicious from legitimate traffic. We apply a constraint-based forwarding to malicious traffic. The goal is to discard enough attack traffic to sustain network stability while preserving legitimate traffic. It is not only aware of attack details but also network resource, especially available bandwidth. Following that network operators do not have equal visibility on their network, we also study the impact of operational constraints on the efficiency of a commonly recommended countermeasure, namely blacklist filtering. The operational criteria are the level of information about the attack and about the traffic inside the network. We then formulate scenario which operators can identify with. We demonstrate that the blacklist generation algorithm should be carefully chosen to fit the operator context while maximizing the filtering efficiency