29
5

Improved Distributed Fractional Coloring Algorithms

Abstract

We prove new bounds on the distributed fractional coloring problem in the LOCAL model. Fractional cc-colorings can be understood as multicolorings as follows. For some natural numbers pp and qq such that p/qcp/q\leq c, each node vv is assigned a set of at least qq colors from {1,,p}\{1,\dots,p\} such that adjacent nodes are assigned disjoint sets of colors. The minimum cc for which a fractional cc-coloring of a graph GG exists is called the fractional chromatic number χf(G)\chi_f(G) of GG. Recently, [Bousquet, Esperet, and Pirot; SIROCCO '21] showed that for any constant ϵ>0\epsilon>0, a fractional (Δ+ϵ)(\Delta+\epsilon)-coloring can be computed in ΔO(Δ)+O(Δlogn)\Delta^{O(\Delta)} + O(\Delta\cdot\log^* n) rounds. We show that such a coloring can be computed in only O(log2Δ)O(\log^2 \Delta) rounds, without any dependency on nn. We further show that in O(lognϵ)O\big(\frac{\log n}{\epsilon}\big) rounds, it is possible to compute a fractional (1+ϵ)χf(G)(1+\epsilon)\chi_f(G)-coloring, even if the fractional chromatic number χf(G)\chi_f(G) is not known. That is, this problem can be approximated arbitrarily well by an efficient algorithm in the LOCAL model. For the standard coloring problem, it is only known that an O(lognloglogn)O\big(\frac{\log n}{\log\log n}\big)-approximation can be computed in polylogarithmic time in the LOCAL model. We also show that our distributed fractional coloring approximation algorithm is best possible. We show that in trees, which have fractional chromatic number 22, computing a fractional (2+ϵ)(2+\epsilon)-coloring requires at least Ω(lognϵ)\Omega\big(\frac{\log n}{\epsilon}\big) rounds. We finally study fractional colorings of regular grids. In [Bousquet, Esperet, and Pirot; SIROCCO '21], it is shown that in regular grids of bounded dimension, a fractional (2+ϵ)(2+\epsilon)-coloring can be computed in time O(logn)O(\log^* n). We show that such a coloring can even be computed in O(1)O(1) rounds in the LOCAL model.

View on arXiv
Comments on this paper