Dice Question Streamline Icon: https://streamlinehq.com

Largest exponent for free-set size in planar graphs

Determine the supremum exponent α such that every n-vertex planar graph contains a free set—i.e., a vertex subset S that can be mapped to any set P of |S| points in the plane while admitting a straight-line crossing-free drawing—of size Ω(n^α). Ascertain whether α equals the shortness exponent σ for cubic triconnected planar graphs by proving or refuting that every triangulation G has a free set of size Ω(c(G^*)), where c(G^*) is the circumference of the dual G^*.

Information Square Streamline Icon: https://streamlinehq.com

Background

This survey connects the size of largest free sets in triangulations to the circumference of their dual cubic triconnected planar graphs. Known upper bounds on dual circumference (shortness exponents) yield upper bounds on free-set size, while various constructive methods yield lower bounds in subclasses.

Currently, the best general bounds imply 1/2 ≤ α ≤ 0.9859, with the upper bound coming from Grünbaum–Walther’s shortness exponent for cubic triconnected planar graphs. Establishing the exact α would unify the combinatorial and geometric aspects of free sets with extremal cycle-length behavior in duals.

References

We conclude with a list of open problems: What is the largest value $\alpha$ such that every $n$-vertex planar graph has a free set of size $\Omega(n{\alpha})$? Currently, we know that $1/2\le \alpha \le 0.9859$. Is $\alpha$ equal to the shortness exponent $\sigma$ for cubic triconnected planar graphs? In other words, does every triangulation $G$ contain a free set of size $\Omega(c(G*))$?

Free Sets in Planar Graphs: History and Applications (2403.17090 - Dujmović et al., 25 Mar 2024) in Section Open Problems (enumerated item 1)