Table of Contents
Fetching ...

An NP-hardness result for the colored constrained maximum 2-edge-colorable subgraph problem in bipartite graphs

Vahan Mkrtchyan

TL;DR

The paper addresses the maximum $k$-edge-colorable subgraph problem, highlighting a dichotomy: it is NP-hard in general/cubic graphs but solvable in polynomial time for bipartite graphs via a max-flow construction. It introduces a vertex-color-constraint variant $ u_2^W(G)$ and proves NP-hardness for connected bipartite graphs with maximum degree $3$ through a reduction from Max $2$-SAT using the KM2008 construction $G_I$, establishing a key relationship $ u_2^W(G_I)= u(G_I)+L(G_I)$. This reduction ties the constrained colorings to special maximum matching structures, demonstrating paraNP-hardness in $k$ and outlining avenues for ETH-based lower bounds and potential exact algorithms. The work links matching theory, flow techniques, and SAT-based reductions to map the boundary between tractable and intractable instances in bipartite graph settings. Overall, it clarifies how vertex-wise color constraints fundamentally alter complexity, even in otherwise tractable bipartite graphs.

Abstract

In this paper, we consider the maximum $k$-edge-colorable subgraph problem. In this problem we are given a graph $G$ and a positive integer $k$, the goal is to take $k$ matchings of $G$ such that their union contains maximum number of edges. This problem is NP-hard in cubic graphs, and polynomial-time solvable in bipartite graphs as we observe in our paper. We present an NP-hardness result for a version of this problem where we have color constraints on vertices. In fact, we show that this version is NP-hard already in bipartite graphs of maximum degree three. In order to achieve the result, we establish a connection between our problem and the problem of construction of special maximum matchings considered in the Master thesis of the author and defended back in 2003.

An NP-hardness result for the colored constrained maximum 2-edge-colorable subgraph problem in bipartite graphs

TL;DR

The paper addresses the maximum -edge-colorable subgraph problem, highlighting a dichotomy: it is NP-hard in general/cubic graphs but solvable in polynomial time for bipartite graphs via a max-flow construction. It introduces a vertex-color-constraint variant and proves NP-hardness for connected bipartite graphs with maximum degree through a reduction from Max -SAT using the KM2008 construction , establishing a key relationship . This reduction ties the constrained colorings to special maximum matching structures, demonstrating paraNP-hardness in and outlining avenues for ETH-based lower bounds and potential exact algorithms. The work links matching theory, flow techniques, and SAT-based reductions to map the boundary between tractable and intractable instances in bipartite graph settings. Overall, it clarifies how vertex-wise color constraints fundamentally alter complexity, even in otherwise tractable bipartite graphs.

Abstract

In this paper, we consider the maximum -edge-colorable subgraph problem. In this problem we are given a graph and a positive integer , the goal is to take matchings of such that their union contains maximum number of edges. This problem is NP-hard in cubic graphs, and polynomial-time solvable in bipartite graphs as we observe in our paper. We present an NP-hardness result for a version of this problem where we have color constraints on vertices. In fact, we show that this version is NP-hard already in bipartite graphs of maximum degree three. In order to achieve the result, we establish a connection between our problem and the problem of construction of special maximum matchings considered in the Master thesis of the author and defended back in 2003.
Paper Structure (3 sections, 13 equations, 10 figures)

This paper contains 3 sections, 13 equations, 10 figures.

Figures (10)

  • Figure 1: A graph in which the maximum matching is not a subset of a maximum $2$-edge-colorable subgraph.
  • Figure 2: The network $H$ obtained from the bipartite graph $G$. The source $s$ of $H$ is joined to every vertex of $A$ with an arc of capacity $k$. Every edge of $G$ is replaced with an arc of capacity $1$. Every vertex of $B$ is joined to the sink $t$ with an arc of capacity $k$.
  • Figure 3: In the path of length four, $\ell(G)=1$ and $L(G)=2$.
  • Figure 4: The gadget corresponding to the variable $x_i$ and the clause $C_j$.
  • Figure 5: The gadget corresponding to the literal $\overline{x}_i$ and clause $C_j$.
  • ...and 5 more figures

Theorems & Definitions (4)

  • proof
  • proof
  • proof
  • proof