The closed neighbourhood of a vertex in a graph is the vertex together with the set of adjacent vertices. A di®erentiating-dominating set, or identifying code, is a collection of vertices whose intersection with the closed neighbour- hoods of each vertex is distinct and nonempty. A di®erentiating-dominating set in a graph serves to uniquely identify all the vertices in the graph. Chapter 1 begins with the necessary de¯nitions and background results and provides motivation for the following chapters. Chapter 1 includes a summary of the lower identi¯cation parameters, °L and °d. Chapter 2 de- ¯nes co-distinguishable graphs and determines bounds on the number of edges in graphs which are distinguishable and co-distinguishable while Chap- ter 3 describes the maximum number of vertices needed in order to identify vertices in a graph, and includes some Nordhaus-Gaddum type results for the sum and product of the di®erentiating-domination number of a graph and its complement. Chapter 4 explores criticality, in which any minor modi¯cation in the edge or vertex set of a graph causes the di®erentiating-domination number to change. Chapter 5 extends the identi¯cation parameters to allow for orientations of the graphs in question and considers the question of when adding orientation helps reduce the value of the identi¯cation parameter. We conclude with a survey of complexity results in Chapter 6 and a collection of interesting new research directions in Chapter 7. / Mathematical Sciences / PhD (Mathematics)

The closed neighbourhood of a vertex in a graph is the vertex together with the set of adjacent vertices. A di®erentiating-dominating set, or identifying code, is a collection of vertices whose intersection with the closed neighbour- hoods of each vertex is distinct and nonempty. A di®erentiating-dominating set in a graph serves to uniquely identify all the vertices in the graph. Chapter 1 begins with the necessary de¯nitions and background results and provides motivation for the following chapters. Chapter 1 includes a summary of the lower identi¯cation parameters, °L and °d. Chapter 2 de- ¯nes co-distinguishable graphs and determines bounds on the number of edges in graphs which are distinguishable and co-distinguishable while Chap- ter 3 describes the maximum number of vertices needed in order to identify vertices in a graph, and includes some Nordhaus-Gaddum type results for the sum and product of the di®erentiating-domination number of a graph and its complement. Chapter 4 explores criticality, in which any minor modi¯cation in the edge or vertex set of a graph causes the di®erentiating-domination number to change. Chapter 5 extends the identi¯cation parameters to allow for orientations of the graphs in question and considers the question of when adding orientation helps reduce the value of the identi¯cation parameter. We conclude with a survey of complexity results in Chapter 6 and a collection of interesting new research directions in Chapter 7. / Mathematical Sciences / PhD (Mathematics)

Un code identifiant est un ensemble de sommets d'un graphe tel que, d'une part, chaque sommet hors du code a un voisin dans le code (propriété de domination) et, d'autre part, tous les sommets ont un voisinage distinct à l'intérieur du code (propriété de séparation). Dans cette thèse, nous nous intéressons à des aspects combinatoires et algorithmiques relatifs aux codes identifiants.Pour la partie combinatoire, nous étudions tout d'abord des questions extrémales en donnant une caractérisation complète des graphes non-orientés finis ayant comme taille minimum de code identifiant leur ordre moins un. Nous caractérisons également les graphes dirigés finis, les graphes non-orientés infinis et les graphes orientés infinis ayant pour seul code identifiant leur ensemble de sommets. Ces résultats répondent à des questions ouvertes précédemment étudiées dans la littérature.Puis, nous étudions la relation entre la taille minimum d'un code identifiant et le degré maximum d'un graphe, en particulier en donnant divers majorants pour ce paramètre en fonction de l'ordre et du degré maximum. Ces majorants sont obtenus via deux techniques. L'une est basée sur la construction d'ensembles indépendants satisfaisant certaines propriétés, et l'autre utilise la combinaison de deux outils de la méthode probabiliste : le lemme local de Lovasz et une borne de Chernoff. Nous donnons également des constructions de familles de graphes en relation avec ce type de majorants, et nous conjecturons que ces constructions sont optimales à une constante additive près.Nous présentons également de nouveaux minorants et majorants pour la cardinalité minimum d'un code identifiant dans des classes de graphes particulières. Nous étudions les graphes de maille au moins 5 et de degré minimum donné en montrant que la combinaison de ces deux paramètres influe fortement sur la taille minimum d'un code identifiant. Nous appliquons ensuite ces résultats aux graphes réguliers aléatoires. Puis, nous donnons des minorants pour la taille d'un code identifiant des graphes d'intervalles et des graphes d'intervalles unitaires. Enfin, nous donnons divers minorants et majorants pour cette quantité lorsque l'on se restreint aux graphes adjoints. Cette dernière question est abordée via la notion nouvelle de codes arête-identifiants.Pour la partie algorithmique, il est connu que le problème de décision associés à la notion de code identifiant est NP-complet même pour des classes de graphes restreintes. Nous étendons ces résultats à d'autres classes de graphes telles que celles des graphes split, des co-bipartis, des adjoints ou d'intervalles. Pour cela nous proposons des réductions polynomiales depuis divers problèmes algorithmiques classiques. Ces résultats montrent que dans beaucoup de classes de graphes, le problème des codes identifiants est algorithmiquement plus difficile que des problèms liés (tel que le problème des ensembles dominants).Par ailleurs, nous complétons les connaissances relatives à l'approximabilité du problème d'optimisation associé aux codes identifiants. Nous étendons le résultat connu de NP-difficulté pour l'approximation de ce problème avec un facteur sous-logarithmique (en fonction de la taille du graphe instance) aux graphes bipartis, split et co-bipartis, respectivement. Nous étendons également le résultat connu d'APX-complétude pour les graphes de degré maximum donné à une sous-classe des graphes split, aux graphes bipartis de degré maximum 4 et aux graphes adjoints. Enfin, nous montrons l'existence d'un algorithme de type PTAS pour les graphes d'intervalles unitaires. / An identifying code is a set of vertices of a graph such that, on the one hand, each vertex out of the code has a neighbour in the code (domination property), and, on the other hand, all vertices have a distinct neighbourhood within the code (separation property). In this thesis, we investigate combinatorial and algorithmic aspects of identifying codes.For the combinatorial part, we first study extremal questions by giving a complete characterization of all finite undirected graphs having their order minus one as minimum size of an identifying code. We also characterize finite directed graphs, infinite undirected graphs and infinite oriented graphs having their whole vertex set as unique identifying code. These results answer open questions that were previously studied in the literature.We then study the relationship between the minimum size of an identifying code and the maximum degree of a graph. In particular, we give several upper bounds for this parameter as a function of the order and the maximum degree. These bounds are obtained using two techniques. The first one consists in the construction of independent sets satisfying certain properties, and the second one is the combination of two tools from the probabilistic method: the Lovasz local lemma and a Chernoff bound. We also provide constructions of graph families related to this type of upper bounds, and we conjecture that they are optimal up to an additive constant.We also present new lower and upper bounds for the minimum cardinality of an identifying code in specific graph classes. We study graphs of girth at least 5 and of given minimum degree by showing that the combination of these two parameters has a strong influence on the minimum size of an identifying code. We apply these results to random regular graphs. Then, we give lower bounds on the size of a minimum identifying code of interval and unit interval graphs. Finally, we prove several lower and upper bounds for this parameter when considering line graphs. The latter question is tackled using the new notion of an edge-identifying code.For the algorithmic part, it is known that the decision problem associated to the notion of an identifying code is NP-complete, even for restricted graph classes. We extend the known results to other classes such as split graphs, co-bipartite graphs, line graphs or interval graphs. To this end, we propose polynomial-time reductions from several classical hard algorithmic problems. These results show that in many graph classes, the identifying code problem is computationally more difficult than related problems (such as the dominating set problem).Furthermore, we extend the knowledge of the approximability of the optimization problem associated to identifying codes. We extend the known result of NP-hardness of approximating this problem within a sub-logarithmic factor (as a function of the instance graph) to bipartite, split and co-bipartite graphs, respectively. We also extendthe known result of its APX-hardness for graphs of given maximum degree to a subclass of split graphs, bipartite graphs of maximum degree 4 and line graphs. Finally, we show the existence of a PTAS algorithm for unit interval graphs.