Core Concepts
The paper presents efficient constructions of non-Steiner tree covers and spanners for planar domains, including polygonal domains, polyhedral terrains, and planar metrics. It identifies a surprising threshold phenomenon around stretch 2, where the number of trees in the non-Steiner tree cover transitions sharply. The paper also provides improved constructions of Steiner spanners for planar domains.
Abstract
The paper studies spanners in planar domains, including polygonal domains, polyhedral terrains, and planar metrics. Previous work showed that for any constant ε ∈ (0,1), one could construct a (2 + ε)-spanner with O(n log(n)) edges, and there is a lower bound of Ω(n^2) edges for any (2 - ε)-spanner. The main open question is whether a linear number of edges suffices and the stretch can be reduced to 2.
The paper resolves this problem by showing that:
- For stretch 2, one needs Ω(n log n) edges. This is the first super-linear lower bound for stretch 2.
- For stretch 2 + ε for any fixed ε ∈ (0,1), O(n) edges are sufficient.
To achieve these results, the paper introduces the problem of constructing non-Steiner tree covers for metrics. It identifies a threshold phenomenon around stretch 2:
- For stretch 2 - ε, Θ(n) trees are necessary and sufficient.
- For stretch 2, Θ(log n) trees are necessary and sufficient.
- For stretch 2 + ε, a constant number of trees suffice.
The paper then uses these non-Steiner tree cover results to construct spanners in planar domains. Specifically:
- For Steiner spanners, the paper constructs a (1 + ε)-spanner with O((n/ε) · log(ε^-1 α(n)) · log ε^-1) edges, where α(n) is the inverse Ackermann function. This improves upon previous results.
- The paper also provides lower bounds for related problems, such as reliable spanners and locality-sensitive orderings, using the non-Steiner tree cover technique.