Fixed-Parameter Complexity of Minimum Profile Problems

TL;DR

This paper investigates the fixed-parameter complexity of the minimum profile problem in graphs, showing that deciding if the profile exceeds a certain bound is fixed-parameter tractable via kernelization.

Contribution

The paper proves fixed-parameter tractability for the problem of determining if a graph's profile is within a certain bound, using a reduction to a linear-size problem kernel.

Findings

Deciding profile ≤ |V|-1+k is fixed-parameter tractable.

Established a linear-size problem kernel for the decision problem.

Provides a new approach to analyze graph profile problems.

Abstract

Let $G=(V,E)$ be a graph. An ordering of $G$ is a bijection $\alpha: V\dom \{1,2,..., |V|\}.$ For a vertex $v$ in $G$, its closed neighborhood is $N[v]=\{u\in V: uv\in E\}\cup \{v\}.$ The profile of an ordering $\alpha$ of $G$ is $\prf_{\alpha}(G)=\sum_{v\in V}(\alpha(v)-\min\{\alpha(u): u\in N[v]\}).$ The profile $\prf(G)$ of $G$ is the minimum of $\prf_{\alpha}(G)$ over all orderings $\alpha$ of $G$. It is well-known that $\prf(G)$ is the minimum number of edges in an interval graph $H$ that contains $G$ is a subgraph. Since $|V|-1$ is a tight lower bound for the profile of connected graphs $G=(V,E)$, the parametrization above the guaranteed value $|V|-1$ is of particular interest. We show that deciding whether the profile of a connected graph $G=(V,E)$ is at most $|V|-1+k$ is fixed-parameter tractable with respect to the parameter $k$. We achieve this result by reduction to a problem kernel of linear size.