22.08.2013 Views

Edge-connectivity of undirected and directed hypergraphs

Edge-connectivity of undirected and directed hypergraphs

Edge-connectivity of undirected and directed hypergraphs

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

12 Chapter 1. Introduction <strong>and</strong> preliminaries<br />

a <strong>directed</strong> graph D = (V, A) whose underlying graph is G.<br />

A set function is a function p : 2 V → Q ∪ {−∞, +∞} (sometimes the set function<br />

is defined only on a given family <strong>of</strong> subsets <strong>of</strong> V ). It is assumed that p(∅) = 0, unless<br />

otherwise stated. For a function m : V → Q <strong>and</strong> a set X ⊆ V , we use the notation<br />

m(X) := <br />

m(v).<br />

v∈X<br />

A graph G is said to cover a set function p if dG(X) ≥ p(X) for every X ⊆ V . Anal-<br />

ogously, a digraph D covers p if ϱD(X) ≥ p(X) for every X ⊆ V . Given a function<br />

m : V → Z+, the graph G satisfies the degree-specification m if dG(v) = m(v) for every<br />

v ∈ V . If we have an in-degree specification mi : V → Z+ <strong>and</strong> an out-degree specifica-<br />

tion mo : V → Z+, the digraph D satisfies the degree-specifications if ϱD(v) = mi(v) <strong>and</strong><br />

δD(v) = mo(v) for every v ∈ V .<br />

In optimization problems we <strong>of</strong>ten consider a cost function c : E → Q on the edges <strong>of</strong><br />

the graph (G = V, E) or c : A → Q on the edges <strong>of</strong> a digraph D = (V, A). A cost function<br />

on E is said to be node induced if c(uv) = c ′ (u) + c ′ (v) where c ′ : V → Q is a cost function<br />

on the nodes. Similarly, a cost function c on A is node induced if c(uv) = c ′ o(u) + c ′ i(v)<br />

where c ′ i : V → Q <strong>and</strong> c ′ o : V → Q are cost functions on the nodes.<br />

1.2.2 The notion <strong>of</strong> edge-<strong>connectivity</strong><br />

For two nodes s ∈ V <strong>and</strong> t ∈ V , a set X ⊆ V is called an st-set if s /∈ X <strong>and</strong> t ∈ X.<br />

Description <strong>of</strong> <strong>un<strong>directed</strong></strong> <strong>and</strong> <strong>directed</strong> edge-<strong>connectivity</strong> is based on the following versions<br />

<strong>of</strong> Menger’s Theorem (see e.g. [21]):<br />

Theorem 1.1 (Menger). Let G = (V, E) be a graph, <strong>and</strong> s, t ∈ V distinct nodes. The<br />

maximum number <strong>of</strong> edge-disjoint paths between s <strong>and</strong> t is<br />

min{dG(X) : X ⊆ V is an st-set}.<br />

Theorem 1.2 (Menger). Let D = (V, A) be a digraph, <strong>and</strong> s, t ∈ V distinct nodes. The<br />

maximum number <strong>of</strong> edge-disjoint <strong>directed</strong> paths from s to t is<br />

min{ϱD(X) : X ⊆ V is an st-set}.<br />

We use the notation λG(s, t) for the maximum number <strong>of</strong> edge-disjoint paths between s<br />

<strong>and</strong> t in G, <strong>and</strong> λD(s, t) for the maximum number <strong>of</strong> edge-disjoint <strong>directed</strong> paths from s<br />

to t in D. These values are called the local edge-<strong>connectivity</strong> between s <strong>and</strong> t (from s to t).<br />

The fact that the global edge-<strong>connectivity</strong> <strong>of</strong> a graph is high can be formulated in different<br />

ways:

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!