A perfect Italian dominating function on a graph G is a function f:V(G)→{0,1,2} satisfying the condition that for every vertex u with f(u)=0, the total weight of f assigned to the neighbors of u is exactly two. The weight of a perfect Italian dominating function is the sum of the weights of the vertices. The perfect Italian domination number of G, denoted γ Ip (G), is the minimum weight of a perfect Italian dominating function of G. We show that if G is a tree on n≥3 vertices, then γ Ip (G)≤[Formula presented]n, and for each positive integer n≡0(mod5) there exists a tree of order n for which equality holds in the bound.

For a graph G=(V,E), we consider placing a variable number of pebbles on the vertices of V. A pebbling move consists of deleting two pebbles from a vertex u∈V and placing one pebble on a vertex v adjacent to u. We seek an initial placement of a minimum total number of pebbles on the vertices in V, so that no vertex receives more than some positive integer t pebbles and for any given vertex v∈V, it is possible, by a sequence of pebbling moves, to move at least one pebble to v. We relate this minimum number of pebbles to several other well-studied parameters of a graph G, including the domination number, the optimal pebbling number, and the Roman domination number of G.

In this thesis, we introduce Roman domination cover rubbling as an extension of domination cover rubbling. We define a parameter on a graph $G$ called the \textit{Roman domination cover rubbling number}, denoted $\rho_{R}(G)$, as the smallest number of pebbles, so that from any initial configuration of those pebbles on $G$, it is possible to obtain a configuration which is Roman dominating after some sequence of pebbling and rubbling moves. We begin by characterizing graphs $G$ having small $\rho_{R}(G)$ value. Among other things, we also obtain the Roman domination cover rubbling number for paths and give an upper bound for the Roman domination cover rubbling number of a tree.

The complementary prism GG of a graph G is formed from the disjoint union of G and its complement G by adding the edges of a perfect match- ing between the corresponding vertices of G and G. A Roman dominating function on a graph G = (V,E) is a labeling f : V(G) → {0,1,2} such that every vertex with label 0 is adjacent to a vertex with label 2. The Roman domination number γR(G) of G is the minimum f(V ) = Σv∈V f(v) over all such functions of G. We study the Roman domination number of complementary prisms. Our main results show that γR(GG) takes on a limited number of values in terms of the domination number of GG and the Roman domination numbers of G and G.

An Italian dominating function on a graph G with vertex set V(G) is a function f: V(G) → { 0 , 1 , 2 } having the property that for every vertex v with f(v) = 0 , at least two neighbors of v are assigned 1 under f or at least one neighbor of v is assigned 2 under f. The weight of an Italian dominating function f is the sum of the values assigned to all the vertices under f. The Italian domination number of G, denoted by γI(G) , is the minimum weight of an Italian dominating of G. It is known that if G is a connected graph of order n≥ 3 , then γI(G)≤34n. Further, if G has minimum degree at least 2, then γI(G)≤23n. In this paper, we characterize the connected graphs achieving equality in these bounds. In addition, we prove Nordhaus–Gaddum inequalities for the Italian domination number.

De nombreux problèmes algorithmiques sont « difficiles », dans le sens où on ne sait pas les résoudre en temps polynomial par rapport à la taille de l’entrée, soit parce qu’ils sont NP-difficiles, soit, pour certains problèmes d’énumération, à cause du nombre exponentiel d'objets à énumérer. Depuis une quinzaine d’années on trouve un intérêt grandissant dans la littérature pour la conception d'algorithmes exacts sophistiqués afin de les résoudre le plus efficacement possible. Dans le cadre de cette thèse, nous nous intéressons à la conception d'algorithmes exacts exponentiels autour de trois problèmes difficiles. Nous étudions tout d'abord le problème d'optimisation Ensemble Connexe Tropical pour lequel nous décrivons un algorithme afin de le résoudre en général, puis un algorithme de branchement plus rapide pour le résoudre sur les arbres, ce problème restant difficile même dans ce cas. Nous nous intéressons ensuite au problème d'énumération Ensembles Dominants Minimaux, pour lequel nous donnons des algorithmes résolvant ce problème dans les graphes splits, cobipartis, ainsi que dans les graphes d'intervalles. Nous déduisons des bornes supérieures sur le nombre d'ensembles dominants minimaux admis par de tels graphes. La dernière étude de cette thèse concerne le problème d'optimisation Domination Romaine Faible dans lequel, étant donné un graphe nous cherchons à construire une fonction de pondération selon certaines propriétés. Le problème est NP-difficile en général, mais nous donnons un algorithme glouton linéaire calculant une telle fonction pour les graphes d'intervalles. / Many algorithmic problems are « hard », in the sense of we do not know how to solve them in polynomialtime, either because they are NP-hard, or, for some enumeration problems, because the number of objectsto be produced is exponential. During the last fifteen years there was a growing interest in the design of exact algorithms to solve such problems as efficiently as possible. In the context of this thesis, we focus on the design of exponential exact algorithms for three hard problems. First, we study the optimisation problem Tropical Connected Set for which we describe an algorithm to solve it in the general case, then a faster branch-and-reduce algorithm to solve it on trees; the problem remains difficult even in this case. Secondly we focus on the Minimal Dominating Sets enumeration problem, for which we give algorithms to solve it on split, cobipartite and intervals graphs. As a byproduct, we establish upper bounds on the number of minimal dominating sets in such graphs. The last focus of this thesis concerns the Weak Roman Domination optimisation problem for which, given a graph, the goal is to build a weight function under some properties. The problem is NP-hard in general, but we give a linear greedy algorithm which computes such a function on interval graphs.