Minimum dilation tree approximation problem

Given a finite point set P⊂R2P\subset\mathbb{R}^2 with ∣P∣=n|P|=n, let TT range over spanning trees on PP, with each edge weighted by its Euclidean length. Define the dilation of TT by δ(T)=max⁡p≠q∈PdT(p,q)∥p−q∥2\delta(T)=\max_{p\ne q\in P}\frac{d_T(p,q)}{\lVert p-q\rVert_2}, where dTd_T is the shortest-path distance in TT, and let OPT⁡(P)=min⁡Tδ(T)\operatorname{OPT}(P)=\min_T\delta(T). Does there exist a constant ε>0\varepsilon>0 and a polynomial-time algorithm that, for every such PP, outputs a spanning tree TT satisfying δ(T)≤O(n1−ε)OPT⁡(P)\delta(T)\le O(n^{1-\varepsilon})\operatorname{OPT}(P)?

References

Progress summary

Refreshed
Claimed progress

A new paper claims the first general improvement over the linear approximation guarantee, but the result has not been independently verified.

The problem asks for an approximation to the minimum-dilation tree problem, posed as an open question by Eppstein. Before this development, the best general guarantee was linear in the number of points.

Known results

  • The minimum spanning tree gives an O(n)O(n) approximation; no o(n)o(n) guarantee was known (2007).
  • The exact Euclidean decision problem is NP\mathrm{NP}-hard for planar point sets (2007).
  • In 2024, the approximation question was explicitly restated as open, including whether an O(n1−ε)O(n^{1-\varepsilon}) approximation exists for some ε>0\varepsilon>0.

September 8, 2026 claimed sublinear approximation

A paper titled A Sublinear Approximation Algorithm for Minimum Dilation Trees in the Plane claims the first asymptotically sublinear general-purpose approximation for planar Euclidean point sets. The claim is reported but not independently verified; hardness and near-optimality remain unsettled.

Current status (as of September 2026): The linear baseline and prior hardness results are established, while a claimed sublinear planar approximation is new and remains unverified.

Sources

Solutions 0

No solutions have been posted yet.