Covering random graphs by monochromatic components
Many problems concerning vertex partitioning and covering in graphs
are still open, even though they have been present in the
literature for more than 50 years. In this work, we are interested
in determining the smallest number $t$ of monochromatic components
needed to cover all the vertices of a binomial random graph,
considering any edge coloring with $r=3$ colors, a property we
denote by $tc_r(G(n,p) )) \leq t$. In the case of $G(n,p)$, in
which the edge density varies as a function of $p=p(n)$, the aim of
this research was to investigate the value of $p$ which
characterizes a threshold function for the property $tc_3 (G(n,p))
\leq 3$. Bal and De Biasio proposed this problem in 2017 and, since
then, some papers have brought advances in understanding this
property. By constructing a counterexample, Ebsen, Mota and
Schnitzer determined a lower bound for $tc_r(G(n,p)) \leq t$, which
occurs asymptotically almost surely with $p \ll \left(\frac {\log
n}{n}\right)^{1/4}$. In 2021, Brada\v{c} and Buci\'{c}
determined the upper bound for the same property with the value of
$p \gg \left(\frac{\log n}{n}\right)^{1 /4}$, through a technique
using auxiliary $r$-uniform hypergraphs. Thus, the threshold
function for the property $tc_3(G(n,p)) \leq 3$ was determined with
the value $p = \left(\frac{\log n}{n}\right)^{1/ 4}$. In this work,
we present the advances made over time regarding this problem, as
well as we present in detail several relevant techniques in
Combinatorics, which were used in order to obtain these advances.