No description available.

MPLS, a next generation backbone architecture, can speed up packet forwarding to destination by label switching. However, if there exists no backup LSP when the primary LSP fails, MPLS frames cannot be forwarded to destination. Therefore, fault recovery has become an important research area in MPLS Traffic Engineering. Makam approach and Haskin approach are the most famous two among the previous literatures. Besides, IETF has made strict definitions for MPLS Recovery in RFC 3469 in February, 2003. We propose a Reverse Labeling Scheme to handle fault recovery in this thesis. We establish a virtual reverse LSP along the completely reverse direction of the primary path. When there is a link failure in the primary LSP, LSR will forward packets back to Ingress by virtual reverse LSP instead of using the primary LSP. This idea of building virtual reverse LSP makes Haskin approach practical in implementation. In addition, we save network resources by designing a scheme such that LSR is easier to convert from the primary LSP to the backup LSP. In order to solve the out-of-order packets in Haskin approach, Hundessa adds buffering on every LSR. The buffer can temporarily store the packets once a link failure has been detected. By adopting the basic idea of Hundessa approach, we embed our Reverse Labeling Scheme and implement it on Linux platform. We also make some modifications to solve the buffering problems. Finally, we demonstrate this Reverse Labeling Scheme by several experiments. We not only show the low packet loss rate, but also solve the packet out-of-order problems. The significant decrease of out-of-order packets can further improve the efficiency of TCP flow transmission.

L’objectif de cette thèse est d’attribuer à chaque sommet x d’un graphe G à n sommets une étiquette L(x) de taille compacte O(log n) bits afin de pouvoir :1. construire, à partir des étiquettes d’un ensemble de sommets en panne X C V (G), une structure de donnée S(X)2. décider, à partir de S(X) et des étiquettes L(u) et L(v), si les sommets u et v sont connectés dans le graphe G n X.Nous proposons une solution à ce problème pour la famille des graphes 3-connexes de genre g (via plusieurs résultats intermédiaires).— Les étiquettes sont de taille O(g log n) bits— Le temps de construction de la structure de donnée S(X) est O(Sort([X]; n)).— Le temps de décision est O(log log n). Ce temps est optimal.Nous étendons ce résultat à la famille des graphes excluant un mineur H fixé. Les étiquettes sont ici de taille O(polylog n) bits. / We aim at assigning each vertex x of a n-vertices graph G a compact O(log n)-bit label L(x) in order to :1. construct, from the labels of the vertices of a forbidden set X C V (G), a datastructure S(X)2. decide, from S(X), L(u) and L(v), whether two vertices u and v are connected in G n X.We give a solution to this problem for the family of 3-connected graphs whith bounded genus.— We obtain O(g log n)-bit labels.— S(X) is computed in O(Sort([X]; n)) time.— Connection between vertices is decided in O(log log n) optimal time.We finally extend this result to H-minor-free graphs. This scheme requires O(polylog n)-bit labels.

In a variety of text document editing and processing applications, it is necessary to keep track of the revision history of text documents by recording changes and the metadata of those changes (e.g., user names and modification timestamps). The recent Web 2.0 document editing and processing applications, such as real-time collaborative note taking and wikis, require fine-grained shared access to collaborative text documents as well as efficient retrieval of metadata associated with different parts of collaborative text documents. Current revision control techniques only support coarse-grained shared access and are inefficient to retrieve metadata of changes at the sub-document granularity. In this dissertation, we design and implement partial persistent sequences (PPSs) to support real-time collaborations and manage metadata of changes at fine granularities for collaborative text document editing and processing applications. As a persistent data structure, PPSs have two important features. First, items in the data structure are never removed. We maintain necessary timestamp information to keep track of both inserted and deleted items and use the timestamp information to reconstruct the state of a document at any point in time. Second, PPSs create unique, persistent, and ordered identifiers for items of a document at fine granularities (e.g., a word or a sentence). As a result, we are able to support consistent and fine-grained shared access to collaborative text documents by detecting and resolving editing conflicts based on the revision history as well as to efficiently index and retrieve metadata associated with different parts of collaborative text documents. We demonstrate the capabilities of PPSs through two important problems in collaborative text document editing and processing applications: data consistency control and fine-grained document provenance management. The first problem studies how to detect and resolve editing conflicts in collaborative text document editing systems. We approach this problem in two steps. In the first step, we use PPSs to capture data dependencies between different editing operations and define a consistency model more suitable for real-time collaborative editing systems. In the second step, we extend our work to the entire spectrum of collaborations and adapt transactional techniques to build a flexible framework for the development of various collaborative editing systems. The generality of this framework is demonstrated by its capabilities to specify three different types of collaborations as exemplified in the systems of RCS, MediaWiki, and Google Docs respectively. We precisely specify the programming interfaces of this framework and describe a prototype implementation over Oracle Berkeley DB High Availability, a replicated database management engine. The second problem of fine-grained document provenance management studies how to efficiently index and retrieve fine-grained metadata for different parts of collaborative text documents. We use PPSs to design both disk-economic and computation-efficient techniques to index provenance data for millions of Wikipedia articles. Our approach is disk economic because we only save a few full versions of a document and only keep delta changes between those full versions. Our approach is also computation-efficient because we avoid the necessity of parsing the revision history of collaborative documents to retrieve fine-grained metadata. Compared to MediaWiki, the revision control system for Wikipedia, our system uses less than 10% of disk space and achieves at least an order of magnitude speed-up to retrieve fine-grained metadata for documents with thousands of revisions.

Tous les problèmes définissables en logique du second ordre monadique peuvent être résolus en temps polynomial dans les classes de graphes qui ont une largeur de clique bornée. La largeur de clique est un paramètre de graphe défini de manière algébrique, c'est-à-dire, à partir d'opérations de composition de graphes. La largeur de rang, définie de manière combinatoire, est une notion équivalente à la largeur de clique des graphes non orientés. Nous donnons une caractérisation algébrique de la largeur de rang et nous montrons qu'elle est linéairement bornée par la largeur arborescente. Nous proposons également une notion de largeur de rang pour les graphes orientés et une relation de vertex-minor pour les graphes orientés. Nous montrons que les graphes orientés qui ont une largeur de rang bornée sont caractérisés par une liste finie de graphes orientés à exclure comme vertex-minor. Beaucoup de classes de graphes n'ont pas une largeur de rang bornée, par exemple, les graphes planaires. Nous nous intéressons aux systèmes d'étiquetage dans ces classes de graphes. Un système d'étiquetage pour une propriété P dans un graphe G, consiste à assigner une étiquette, aussi petite que possible, à chaque sommet de telle sorte que l'on puisse vérifier si G satisfait P en n'utilisant que les étiquettes des sommets. Nous montrons que si P est une propriété définissable en logique du premier ordre alors, certaines classes de graphes de largeur de clique localement bornée admettent un système d'étiquetage pour P avec des étiquettes de taille logarithmique. Parmi ces classes on peut citer les classes de graphes de degré borné, les graphes planaires et plus généralement les classes de graphes qui excluent un apex comme mineur et, les graphes d'intervalle unitaire. Si x et y sont deux sommets, X un ensemble de sommets et F un ensemble d'arêtes, nous notons Conn(x,y,X,F) la propriété qui vérifie dans un graphe donné si x et y sont connectés par un chemin, qui ne passe par aucun sommet de X si aucune arête de F. Cette propriété n'est pas définissable en logique du premier ordre. Nous montrons qu'elle admet un système d'étiquetage avec des étiquettes de taille logarithmique dans les graphes planaires. Nous montrons enfin que Conn(x,y,X,0) admet également un système d'étiquetage avec des étiquettes de taille logarithmique dans des classes de graphes qui sont définies comme des combinaisons de graphes qui ont une petite largeur de clique et telles que le graphe d'intersection de ces derniers est planaire et est de degré borné. / Every property definable in onadic second order logic can be checked in polynomial-time on graph classes of bounded clique-width. Clique-width is a graph parameter defined in an algebraical way, i.e., with operations ``concatenating graphs'' and that generalize concatenation of words.Rank-width, defined in a combinatorial way, is equivalent to the clique-width of undirected graphs. We give an algebraic characterization of rank-width and we show that rank-width is linearly bounded in term of tree-width. We also propose a notion of ``rank-width'' of directed graphs and a vertex-minor inclusion for directed graphs. We show that directed graphs of bounded ``rank-width'' are characterized by a finite list of finite directed graphs to exclude as vertex-minor. Many graph classes do not have bounded rank-width, e.g., planar graphs. We are interested in labeling schemes on these graph classes. A labeling scheme for a property P in a graph G consists in assigning a label, as short as possible, to each vertex of G and such that we can verify if G satisfies P by just looking at the labels. We show that every property definable in first order logic admit labeling schemes with labels of logarithmic size on certain graph classes that have bounded local clique-width. Bounded degree graph classes, minor closed classes of graphs that exclude an apex graph as a minor have bounded local clique-width. If x and y are two vertices and X is a subset of the set of vertices and Y is a subset of the set of edges, we let Conn(x,y,X,Y) be the graph property x and y are connected by a path that avoids the vertices in X and the edges in Y. This property is not definable by a first order formula. We show that it admits a labeling scheme with labels of logarithmic size on planar graphs. We also show that Conn(x,y,X,0) admits short labeling schemes with labels of logarithmic size on graph classes that are ``planar gluings'' of graphs of small clique-width and with limited overlaps.