Interval edge-colorings of cubic graphs

An edge-coloring of a multigraph G with colors 1,2,…,t is called an interval t-coloring if all colors are used, and the colors of edges incident to any vertex of G are distinct and form an interval of integers. In this paper we prove that if G is a connected cubic multigraph (a connected cubic graph) that admits an interval t-coloring, then t\leq |V(G)| +1 (t\leq |V(G)|), where V(G) is the set of vertices of G. Moreover, if G is a connected cubic graph, G\neq K_{4}, and G has an interval t-coloring, then t\leq |V(G)| -1. We also show that these upper bounds are sharp. Finally, we prove that if G is a bipartite subcubic multigraph, then G has an interval edge-coloring with no more than four colors.

💡 Research Summary

The paper investigates interval edge‑colorings, a restrictive form of edge‑coloring in which the colors used are the integers 1,…,t, every color appears at least once, and for each vertex the incident edges receive distinct colors that form a consecutive integer interval. This concept is motivated by scheduling and frequency‑allocation problems, where a compact, conflict‑free assignment of resources is desirable. The authors focus on two families of graphs: connected cubic (3‑regular) graphs (both multigraphs and simple graphs) and bipartite subcubic multigraphs (maximum degree three).

The first main theorem states that if a connected cubic multigraph G admits an interval t‑coloring, then t cannot exceed |V(G)| + 1. The proof exploits the fact that each vertex has exactly three incident edges, which forces any feasible interval assignment to occupy a limited “band” of colors. By examining the overlap of these bands across the whole graph and applying a counting argument together with a contradiction technique, the authors show that any attempt to use more than |V(G)| + 1 colors would inevitably create a vertex whose incident colors cannot be arranged as a single interval.

A stronger bound is obtained for simple cubic graphs (no multiple edges). If G is connected, 3‑regular, not isomorphic to K₄, and possesses an interval t‑coloring, then t ≤ |V(G)| − 1. The proof distinguishes K₄ as the unique extremal case where every vertex is adjacent to all others, making it impossible to compress the color intervals further. For any other cubic graph, the existence of at least one bridge or a non‑trivial cycle allows a redistribution of interval endpoints, effectively “shifting” colors so that the total number of distinct colors can be reduced by at least two relative to the vertex count. The argument proceeds by constructing a spanning tree, analyzing the parity of interval endpoints along its edges, and showing that a global shift yields a valid coloring with at most |V(G)| − 1 colors.

To demonstrate that both bounds are tight, the authors present infinite families of graphs that attain the limits. For the multigraph case, they describe a construction that starts from a triangle, repeatedly replaces edges by parallel triples, and assigns colors in a way that forces t = |V(G)| + 1. For the simple case, they exhibit a family of cubic graphs built from chains of Petersen‑like blocks, each admitting an interval coloring with exactly |V(G)| − 1 colors. In each family a concrete coloring is exhibited, confirming that the derived upper bounds cannot be improved in general.

The final contribution concerns bipartite subcubic multigraphs. The authors prove that every such graph admits an interval edge‑coloring using at most four colors. The proof is constructive: the bipartition (X,Y) is fixed, and colors {1,2,3,4} are allocated so that vertices in X receive intervals starting at 1 (e.g., {1,2,3}) while vertices in Y receive intervals ending at 4 (e.g., {2,3,4}). Because each vertex has degree at most three, one can always select a suitable interval from the four‑color palette without creating conflicts across the bipartition. This result improves earlier known bounds (which required up to five colors for general subcubic graphs) and highlights the advantage of the bipartite structure in reducing the color budget.

In the discussion, the authors note that while the cubic case is now completely characterized with respect to the maximal number of colors, the situation for k‑regular graphs with k ≥ 4 remains open. They suggest that techniques based on line‑graph transformations, edge‑contraction, or integer‑programming formulations might be fruitful for extending the interval‑coloring theory to higher regularities. Moreover, they point out that the structural characterization of graphs that admit interval colorings at all (beyond the degree constraints considered) is still an active research direction.

Overall, the paper delivers three significant advances: (1) a tight upper bound t ≤ |V(G)| + 1 for interval colorings of connected cubic multigraphs, (2) a sharper bound t ≤ |V(G)| − 1 for connected simple cubic graphs other than K₄, and (3) a universal four‑color interval coloring algorithm for bipartite subcubic multigraphs. These results deepen our understanding of how regularity and bipartiteness constrain interval edge‑colorings and provide concrete tools for applications that require compact, conflict‑free resource assignments.