On the Complexity of the Minimum-($k,ρ$)-Shortcut Problem

Abstract

We consider the Minimum-$(k,ρ)$-$\mathrm{Shortcut}$ problem ($\min(k,ρ)\text{-}\mathrm{Shortcut}$), where the goal is to find the smallest set of shortcut edges such that every vertex in a given graph can reach its $ρ$ closest vertices using paths of at most $k$ edges. This is a fundamental graph optimization problem used to accelerate parallel shortest path algorithms. It is well-known that the problem is trivially solvable for the cases $k=1$ and $k\geqρ$. While recent work by Leonhardt, Meyer, and Penschuck (ESA 2024) showed that in undirected graphs $\min(k,ρ)\text{-}\mathrm{Shortcut}$ is NP-hard for $k\geq 3$ if $ρ=Θ(n^ε)$, the boundary where the problem transitions from polynomial-time solvable to NP-hard remained open. In this paper, we narrow this gap significantly. We present a simpler and more direct reduction from the Hitting Set problem which establishes that $\min(k,ρ)\text{-}\mathrm{Shortcut}$ is NP-hard for $k\geq2$ and $ρ\geq k+2$ in both directed and undirected graphs. Complementing this, we use the symmetry of the undirected case to show that $ρ=k+1$ is solvable in polynomial time, a regime where the directed version remains a candidate for NP-hardness. Therefore, we obtain an almost complete characterization of the complexity of $\min(k,ρ)\text{-}\mathrm{Shortcut}$, with the sole remaining open case being $ρ= k+1$ in the directed setting.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.