Table of Contents
Fetching ...

Separation Number and Treewidth, Revisited

Hussein Houdrouge, Babak Miraftab, Pat Morin

TL;DR

This work establishes a linear upper bound on the treewidth in terms of the separation number by a constructive proof that avoids brambles, havens, and network-flow machinery. It introduces generalized $W$-balanced separations and uses Menger-type path arguments to iteratively build a tree decomposition, ensuring all bags stay within a constant factor of the separation number. The core technical contribution is a careful analysis (via $W$-sequences and a key bound) that enables a recursive assembly of decompositions on subgraphs around a balanced separation. As a result, the paper provides an explicit constant $c=7915/139<56.943$ with a polynomial-time construction of the decomposition, offering a practical, threshold-based relation between $tw(G)$ and $sn(G)$.

Abstract

We give a constructive proof of the fact that the treewidth of a graph $G$ is bounded by a linear function of the separation number of $G$.

Separation Number and Treewidth, Revisited

TL;DR

This work establishes a linear upper bound on the treewidth in terms of the separation number by a constructive proof that avoids brambles, havens, and network-flow machinery. It introduces generalized -balanced separations and uses Menger-type path arguments to iteratively build a tree decomposition, ensuring all bags stay within a constant factor of the separation number. The core technical contribution is a careful analysis (via -sequences and a key bound) that enables a recursive assembly of decompositions on subgraphs around a balanced separation. As a result, the paper provides an explicit constant with a polynomial-time construction of the decomposition, offering a practical, threshold-based relation between and .

Abstract

We give a constructive proof of the fact that the treewidth of a graph is bounded by a linear function of the separation number of .

Paper Structure

This paper contains 3 sections, 7 theorems, 21 equations.

Key Result

Theorem 1

There exists a constant $c$ such that, for every graph $G$, $\mathop{\mathrm{tw}}\nolimits(G)\le c\cdot \mathop{\mathrm{sn}}\nolimits(G)$.

Theorems & Definitions (19)

  • Theorem 1
  • Theorem 2: robertson.seymour:graph
  • Theorem 3: Menger's Theorem
  • Lemma 4
  • proof
  • Lemma 5
  • proof
  • Lemma 6
  • proof
  • Lemma 7
  • ...and 9 more