An extension on neighbor sum distinguishing total coloring of graphs

TL;DR

This paper introduces a new type of total coloring for graphs called neighbor full sum distinguishing total coloring, and proves the conjecture that its chromatic number is at most 3 for various classes of graphs.

Contribution

It extends the concept of neighbor sum distinguishing total coloring and proves the conjecture for several important classes of graphs.

Findings

The conjecture holds for paths and cycles.

The conjecture holds for 3-regular graphs.

Complete graphs achieve the upper bound.

Abstract

Let $f: V(G)\cup E(G)\rightarrow \{1,2,\dots,k\}$ be a non-proper total $k$-coloring of $G$. Define a weight function on total coloring as $$\phi(x)=f(x)+\sum\limits_{e\ni x}f(e)+\sum\limits_{y\in N(x)}f(y),$$ where $N(x)=\{y\in V(G)|xy\in E(G)\}$. If $\phi(x)\neq \phi(y)$ for any edge $xy\in E(G)$, then $f$ is called a neighbor full sum distinguishing total $k$-coloring of $G$. The smallest value $k$ for which $G$ has such a coloring is called the neighbor full sum distinguishing total chromatic number of $G$ and denoted by fgndi$_{\sum}(G)$. The coloring is an extension of neighbor sum distinguishing non-proper total coloring. In this paper we conjecture that fgndi$_{\sum}(G)\leq 3$ for any connected graph $G$ of order at least three. We prove that the conjecture is true for (i) paths and cycles; (ii) 3-regular graphs and (iii) stars, complete graphs, trees, hypercubes, bipartite graphs and complete $r$-partite graphs. In particular, complete graphs can achieve the upper bound for the above conjecture.