Polynomially solvable cases for the maximum stable set problem

We describe two classes of graphs for which the stability number can be computed in polynomial time. The first approach is based on an iterative procedure which, at each step, builds from a graph G a new graph G′ which has fewer vertices and has the property that (G′) = (G) − 1. For the second class, it can be decided in polynomial time whether there exists a stable set of given size k.     

On-line algorithms for ordered sets and comparability graphs

We discuss several results concerning on-line algorithms for ordered sets and comparability graphs. The main new result is on the problem of on-line transitive orientation. We view on-line transitive orientation of a comparability graph G as a two-person game. In the ith round of play, 1 i | V(G)|, player A names a graph G_i such that G_i is isomorphic to a subgraph of G, |V(G_i)| = i, and G_i−1 is an induced subgraph of G_i if i> 1. Player B must respond with a transitive orientation of G_i which extends the transitive orientation given to G_i−1 in the previous round of play. Player A wins if and only if player B fails to give a transitive orientation to G_i for some i, 1 i |V(G)|. Our main result shows that player A has at most three winning moves. We also discuss applications to on-line clique covering of comparability graphs, and we mention some open problems.     

Some results on characterizing the edges of connected graphs with a given domination number

A dominating set for a graph G = (V, E) is a subset of vertices V′ V such that for all v ε V − V′ there exists some u ε V′ for which {v, u} ε E. The domination number of G is the size of its smallest dominating set(s). For a given graph G with minimum size dominating set D, let m₁ (G, D) denote the number of edges that have neither endpoint in D, and let m₂ (G, D) denote the number of edges that have at least one endpoint in D. We characterize the possible values that the pair (m₁ (G, D), m₂ (G, D)) can attain for connected graphs having a given domination number.     

Nonchordal positive semidefinite stochastic matrices

Let K_nbe the convex set of n×npositive semidefinite doubly stochastic matrices. If Aε k_n, the graph of A,G(A), is the graph on n vertices with (i,j) an edge if a_ij ≠ 0i≠ j. We are concerned with the extreme points in K_n. In many cases, the rank of Aand G(A) are enough to determine whether A is extreme in K_n. This is true, in particular, if G(A)is a special kind of nonchordal graph, i.e., if no two cycles in G(A)have a common edge.     

Reconstruction of infinite locally finite connected graphs

Let G be an infinite locally finite connected graph. We study the reconstructibility of G in relation to the structure of its end set . We prove that an infinite locally finite connected graph G is reconstructible if there exists a finite family (Ω_i)_0i (n2) of pairwise finitely separable subsets of such that, for all x,y,x′,y′V(G) and every isomorphism f of G−{x,y} onto G−{x′,y′} there is a permutation π of {0,…,n−1} such that for 0i<n. From this theorem we deduce, as particular consequences, that G is reconstructible if it satisfies one of the following properties: (i) G contains no end-respecting subdivision of the dyadic tree and has at least two ends of maximal order; (ii) the set of thick ends or the one of thin ends of G is finite and of cardinality greater than one. We also prove that if almost all vertices of G are cutvertices, then G is reconstructible if it contains a free end or if it has at least a vertex which is not a cutvertex.     

A note on minimum degree conditions for supereulerian graphs

A graph is called supereulerian if it has a spanning closed trail. Let G be a 2-edge-connected graph of order n such that each minimal edge cut SE(G) with |S|3 satisfies the property that each component of G−S has order at least (n−2)/5. We prove that either G is supereulerian or G belongs to one of two classes of exceptional graphs. Our results slightly improve earlier results of Catlin and Li. Furthermore, our main result implies the following strengthening of a theorem of Lai within the class of graphs with minimum degree δ4: If G is a 2-edge-connected graph of order n with δ(G)4 such that for every edge xyE(G) , we have max{d(x),d(y)}(n−2)/5−1, then either G is supereulerian or G belongs to one of two classes of exceptional graphs. We show that the condition δ(G)4 cannot be relaxed.     

Hamiltonian cycles in n-factor-critical graphs

A graph G is said to be n-factor-critical if G−S has a 1-factor for any SV(G) with |S|=n. In this paper, we prove that if G is a 2-connected n-factor-critical graph of order p with , then G is hamiltonian with some exceptions. To extend this theorem, we define a (k,n)-factor-critical graph to be a graph G such that G−S has a k-factor for any SV(G) with |S|=n. We conjecture that if G is a 2-connected (k,n)-factor-critical graph of order p with , then G is hamiltonian with some exceptions. In this paper, we characterize all such graphs that satisfy the assumption, but are not 1-tough. Using this, we verify the conjecture for k2.     

On the minimal reducible bound for outerplanar and planar graphs

Let L be the set of all additive and hereditary properties of graphs. For P₁, P₂ L we define the reducible property R = P₁ P₂ as follows: G P₁P₂ if there is a bipartition (V₁, V₂) of V(G) such that V₁ P₁ and V₂ P₂. For a property P L, a reducible property R is called a minimal reducible bound for P if P R and for each reducible property R′, R′ R → P R′. It is proved that the class of all outerplanar graphs has exactly two minimal reducible bounds in L. Some related problems for planar graphs are discussed.     

The relationship between the threshold dimension of split graphs and various dimensional parameters

Let the coboxicity of a graph G be denoted by cob(G), and the threshold dimension by t(G). For fixed k≥3, determining if cob(G)≥k and t(G)≤k are both NP-complete problems. We show that if G is a comparability graph, then we can determine if cob(G)≤2 in polynomial time. This result shows that it is possible to determine if the interval dimension of a poset equals 2 in polynomial time. If the clique covering number of G is 2, we show that one can determine if t(G)≤2 in polynomial time. Sufficient conditions on G are given for cob(G)≤2 and for t(G)≤2.     

Circulant graphs with det(−A(G))=−deg(G): codeterminants with Kn

Circulant graphs satisfying det(−A(G))=−deg(G) are used to construct arbitrarily large families of graphs with determinant equal to that of the complete graph K_n.     

On the Differential Polynomial of a Graph

We introduce the differential polynomial of a graph. The differential polynomial of a graph G of order n is the polynomial B(G; x) :=∑?(G)k=-nB_k(G) x~(n+k), where B_k(G) denotes the number of vertex subsets of G with differential equal to k. We state some properties of B(G;x) and its coefficients.In particular, we compute the differential polynomial for complete, empty, path, cycle, wheel and double star graphs. We also establish some relationships between B(G; x) and the differential polynomials of graphs which result by removing, adding, and subdividing an edge from G.     

The coordinate representation of a graph and n-universal graph of radius 1

Every graph can be represented as the intersection graph on a family of closed unit cubes in Euclidean space Eⁿ. Cube vertices have integer coordinates. The coordinate matrix, A(G)={v_nk} of a graph G is defined by the set of cube coordinates. The imbedded dimension of a graph, Bp(G), is a number of columns in matrix A(G) such that each of them has at least two distinct elements v_nk≠v_pk. We show that Bp(G)=cub(G) for some graphs, and Bp(G)n−2 for any graph G on n vertices. The coordinate matrix uses to obtain the graph U of radius 1 with 3ⁿ⁻² vertices that contains as an induced subgraph a copy of any graph on n vertices.     

Maximum number of edges in a critically k-connected graph

A k-connected graph G is said to be critically k-connected if G−v is not k-connected for any vV(G). We show that if n,k are integers with k4 and nk+2, and G is a critically k-connected graph of order n, then |E(G)|n(n−1)/2−p(n−k)+p²/2, where p=(n/k)+1 if n/k is an odd integer and p=n/k otherwise. We also characterize extremal graphs.     

The complexity of irredundant sets parameterized by size

An irredundant set of vertices V′V in a graph G=(V,E) has the property that for every vertex uV′, N[V′−{u}] is a proper subset of N[V′]. We investigate the parameterized complexity of determining whether a graph has an irredundant set of size k, where k is the parameter. The interest of this problem is that while most “k-element vertex set” problems are NP-complete, several are known to be fixed-parameter tractable, and others are hard for various levels of the parameterized complexity hierarchy. Complexity classification of vertex set problems in this framework has proved to be both more interesting and more difficult. We prove that the k-element irredundant set problem is complete for W[1], and thus has the same parameterized complexity as the problem of determining whether a graph has a k-clique. We also show that the “parametric dual” problem of determining whether a graph has an irredundant set of size n−k is fixed-parameter tractable.     

Optimally super-edge-connected transitive graphs

Let X=(V,E) be a connected regular graph. X is said to be super-edge-connected if every minimum edge cut of X is a set of edges incident with some vertex. The restricted edge connectivity λ′(X) of X is the minimum number of edges whose removal disconnects X into non-trivial components. A super-edge-connected k-regular graph is said to be optimally super-edge-connected if its restricted edge connectivity attains the maximum 2k−2. In this paper, we define the λ′-atoms of graphs with respect to restricted edge connectivity and show that if X is a k-regular k-edge-connected graph whose λ′-atoms have size at least 3, then any two distinct λ′-atoms are disjoint. Using this property, we characterize the super-edge-connected or optimally super-edge-connected transitive graphs and Cayley graphs. In particular, we classify the optimally super-edge-connected quasiminimal Cayley graphs and Cayley graphs of diameter 2. As a consequence, we show that almost all Cayley graphs are optimally super-edge-connected.     

Connectivity of generalized prisms over G

The problem of building larger graphs with a given graph as an induced subgraph is one which can arise in various applications and in particular can be important when constructing large communications networks from smaller ones. What one can conclude from this paper is that generalized prisms over G may provide an important such construction because the connectivity of the newly created graph is larger than that of the original (connected) graph, regardless of the permutation used.

For a graph G with vertices labeled 1,2,…, n and a permutation in S_n, the generalized prisms over G, (G) (also called a permutation graph), consists of two copies of G, say G_x and G_y, along with the edges (x_i, y_(i), for 1≤i≤n. The purpose of this paper is to examine the connectivity of generalized prisms over G. In particular, upper and lower bounds are found. Also, the connectivity and edge-connectivity are determined for generalized prisms over trees, cycles, wheels, n-cubes, complete graphs, and complete bipartite graphs. Finally, the connectivity of the generalized prism over G, (G), is determined for all in the automorphism group of G.     

Optimal multiple interval assignments in frequency assignment and traffic phasing

In this paper, we consider the optimal assignments of unions of intervals to the vertices of the compatibility graph G, which arises in connection with frequency assignment and traffic phasing problems. It is shown that the optimal multiple interval phasing numbers θ_J≥rx(G) and θ^J≥r_xN(G), are optimal solutions to linear programming problems whose variables correspond to maximal cliques of G. Efficient algorithms are given for determining the first number, θ_J≥rx(G), when G is a chordal graph or a transitively orientable graph.     

The integrity of a cubic graph

Integrity, a measure of network reliability, is defined as

where G is a graph with vertex set V and m(G−S) denotes the order of the largest component of G−S. We prove an upper bound of the following form on the integrity of any cubic graph with n vertices:

Moreover, there exist an infinite family of connected cubic graphs whose integrity satisfies a linear lower bound I(G)>βn for some constant β. We provide a value for β, but it is likely not best possible. To prove the upper bound we first solve the following extremal problem. What is the least number of vertices in a cubic graph whose removal results in an acyclic graph? The solution (with a few minor exceptions) is that n/3 vertices suffice and this is best possible.     

Algorithmic aspects of clique-transversal and clique-independent sets

A minimum clique-transversal set MCT(G) of a graph G=(V,E) is a set SV of minimum cardinality that meets all maximal cliques in G. A maximum clique-independent set MCI(G) of G is a set of maximum number of pairwise vertex-disjoint maximal cliques. We prove that the problem of finding an MCT(G) and an MCI(G) is NP-hard for cocomparability, planar, line and total graphs. As an interesting corollary we obtain that the problem of finding a minimum number of elements of a poset to meet all maximal antichains of the poset remains NP-hard even if the poset has height two, thereby generalizing a result of Duffas et al. (J. Combin. Theory Ser. A 58 (1991) 158–164). We present a polynomial algorithm for the above problems on Helly circular-arc graphs which is the first such algorithm for a class of graphs that is not clique-perfect. We also present polynomial algorithms for the weighted version of the clique-transversal problem on strongly chordal graphs, chordal graphs of bounded clique size, and cographs. The algorithms presented run in linear time for strongly chordal graphs and cographs. These mark the first attempts at the weighted version of the problem.     

The chromatic difference sequence of the Cartesian product of graphs

The chromatic difference sequence cds(G) of a graph G with chromatic number n is defined by cds(G) = (a(1), a(2),…, a(n)) if the sum of a(1), a(2),…, a(t) is the maximum number of vertices in an induced t-colorable subgraph of G for t = 1, 2,…, n. The Cartesian product of two graphs G and H, denoted by G□H, has the vertex set V(G□H = V(G) x V(H) and its edge set is given by (x₁, y₁)(x₂, y₂) ε E(G□H) if either x₁ = x₂ and y₁ y₂ ε E(H) or y₁ = y₂ and x₁x₂ ε E(G).

We obtained four main results: the cds of the product of bipartite graphs, the cds of the product of graphs with cds being nondrop flat and first-drop flat, the non-increasing theorem for powers of graphs and cds of powers of circulant graphs.