[Paper Review] Counting Euclidean embeddings of rigid graphs

Ioannis Z. Emiris, Ioannis Psarros|arXiv (Cornell University)|Feb 6, 2014

Structural Analysis and Optimization10 references3 citations

TL;DR

This paper presents a corrected algebraic geometry approach to upper-bounding the number of planar Euclidean embeddings of minimally rigid (Laman) graphs, using distance geometry and mixed volume computation of polynomial systems derived from Cayley-Menger matrices. It rectifies prior errors in bounds for n=8 and confirms a tight upper bound of 56 embeddings for n=7, with a conjecture that mixed volumes from larger systems may already account for reflections, eliminating the need for doubling the root count.

ABSTRACT

A graph is called (generically) rigid in $\mathbb{R}^d$ if, for any choice of sufficiently generic edge lengths, it can be embedded in $\mathbb{R}^d$ in a finite number of distinct ways, modulo rigid transformations. Here we deal with the problem of determining the maximum number of planar Euclidean embeddings as a function of the number of the vertices. We obtain polynomial systems which totally capture the structure of a given graph, by exploiting distance geometry theory. Consequently, counting the number of Euclidean embeddings of a given rigid graph, reduces to the problem of counting roots of the corresponding polynomial system.

Motivation & Objective

• To determine the maximum number of distinct planar Euclidean embeddings of minimally rigid graphs modulo rigid transformations.
• To correct inaccuracies in prior bounds for Laman graphs with 8 vertices, particularly those based on flawed polynomial systems.
• To establish a reliable method using distance geometry and mixed volume theory to bound the number of embeddings via root counting of polynomial systems.
• To investigate whether mixed volume computations from larger systems inherently account for reflection symmetries, potentially eliminating the need to double root counts.
• To provide tight upper bounds for n ≤ 10, especially confirming the tightness of the bound for n=7.

Proposed method

• Constructs a polynomial system from the Cayley-Menger matrix of a Laman graph, encoding distance constraints via squared edge lengths.
• Uses the mixed volume of the Newton polytopes of the system's equations as an upper bound on the number of isolated complex roots, which corresponds to embeddings modulo rigid transformations.
• Applies Henneberg constructions (H1 and H2 steps) to inductively build Laman graphs and track the effect on embedding counts: H1 steps double the number of embeddings, H2 steps are analyzed via system size and mixed volume.
• Imposes rank and determinant conditions on Cayley-Menger minors to ensure realizability in R², forming well-constrained systems that uniquely determine embeddings up to symmetry.
• Employs distance geometry theory to derive systems of equations whose roots correspond to valid Euclidean embeddings, with reflection symmetry handled via conjecture on system completeness.
• Validates results through mixed volume computation on systems of size (n-3)×(n-3) for n≤7 and (n-2)×(n-2) for n=8, with conjectures on reflection invariance.

Experimental results

Research questions

• RQ1 What is the correct upper bound on the number of planar Euclidean embeddings for Laman graphs with 8 vertices, given prior results were based on flawed systems?
• RQ2 Can mixed volume computations from larger (n-2)×(n-2) systems for n=8 already account for reflection symmetries, eliminating the need to multiply by 2?
• RQ3 Why did the prior bound of 128 for n=8 fail, and what structural property of the systems caused the error?
• RQ4 Is the upper bound of 56 embeddings for 7-vertex Laman graphs tight, and does it hold across all H2-type graphs?
• RQ5 Can a general algebraic method based on distance geometry and mixed volume provide tight, combinatorially meaningful bounds for larger n?

Key findings

• The upper bound for 7-vertex Laman graphs is confirmed as 56, achieved via a mixed volume computation of 28 on a (n-3)×(n-3) system, which is tight and matches known results.
• The prior bound of 128 for 8-vertex Laman graphs is incorrect due to systems that violate the Laman property, invalidating the root count as a valid upper bound.
• For 8-vertex graphs, systems of size (n-2)×(n-2) yield mixed volumes that may already account for reflection, suggesting a conjecture that no doubling is needed, leading to an expected upper bound of 122 instead of 128.
• The method successfully corrects the proof for n=7, which was previously flawed due to unaccounted mirror images, but the final bound remains valid.
• The approach via distance geometry and mixed volume provides a systematic framework for bounding embeddings, with H1 steps doubling the count and H2 steps contributing multiplicatively, though the exact factor remains conjectural.
• For n=9 and n=10, the upper bounds are 512 and 2048 respectively, derived as multiples of the previous bounds, though the paper notes these may not be tight.