[Paper Review] Characteristics of graph braid groups

Ki Hyoung Ko, Hyo Won Park|arXiv (Cornell University)|Jan 13, 2011

Topological and Geometric Data Analysis6 references4 citations

TL;DR

This paper provides explicit formulae for the first homology of graph braid groups in terms of graph-theoretic invariants, proving that the first homology of the n-braid group over a finite graph is torsion-free if and only if the graph is planar. It confirms conjectures on pure 2-braid group homology and establishes that both the 2-braid and pure 2-braid groups over planar graphs have presentations with relators that are commutators of words corresponding to disjoint cycles, resolving a long-standing conjecture from [9].

ABSTRACT

We give formulae for the first homology of the $n$-braid group and the pure 2-braid group over a finite graph in terms of graph theoretic invariants. As immediate consequences, a graph is planar if and only if the first homology of the $n$-braid group over the graph is torsion-free and the conjectures about the first homology of the pure 2-braid groups over graphs in \cite{FH} can be verified. We discover more characteristics of graph braid groups: the $n$-braid group over a planar graph and the pure 2-braid group over any graph have a presentation whose relators are words of commutators, and the 2-braid group and the pure 2-braid group over a planar graph have a presentation whose relators are commutators. The latter was a conjecture in \cite{FS2} and so we propose a similar conjecture for higher braid indices.

Motivation & Objective

• To derive explicit formulae for the first homology of the n-braid group and pure 2-braid group over finite graphs using graph-theoretic invariants.
• To verify long-standing conjectures regarding the first homology of pure 2-braid groups over graphs, particularly concerning torsion and Betti numbers.
• To establish that the 2-braid group and pure 2-braid group over planar graphs admit presentations with relators that are commutators of disjoint cycles.
• To propose and support a conjecture on the structure of higher braid index groups over planar graphs, excluding subgraphs isomorphic to Θ₄.

Proposed method

• Utilizes discrete configuration spaces $D_n\Gamma$ and $UD_n\Gamma$ as cubical complexes deformation retracts of the continuous configuration spaces $C_n\Gamma$ and $UC_n\Gamma$.
• Applies homological algebra techniques, including the use of exact sequences involving $H_2(D_2\Gamma)$, $H_2(\Gamma \times \Gamma)$, and relative homology to analyze 2-dimensional homology.
• Employs Tietze transformations on group presentations to eliminate generator-relation pairs, corresponding to collapsing canceling 1- and 2-cells in Morse complexes.
• Uses the Mayer-Vietoris sequence and Euler characteristic arguments to relate the number of 2-cells in the Morse complex to the second Betti number.
• Leverages the fact that planar graphs admit disjoint cycle pairs whose boundaries generate $H_2(UD_2\Gamma)$, leading to commutator relators.
• Applies results from [3] and [12] to connect Betti numbers of configuration spaces to those of the underlying graph.

Experimental results

Research questions

• RQ1 When is the first homology of the n-braid group over a finite graph torsion-free?
• RQ2 What is the precise formula for the first homology of the pure 2-braid group over a finite graph in terms of graph invariants?
• RQ3 Are the 2-braid and pure 2-braid groups over planar graphs simple-commutator-related with relators being commutators of disjoint cycles?
• RQ4 Under what conditions is the n-braid group over a planar graph simple-commutator-related, particularly for $n \geq 3$?
• RQ5 Does the presence of a $\Theta_4$ subgraph obstruct the simple-commutator-related structure in higher braid groups?

Key findings

• The first homology of the n-braid group over a finite graph is torsion-free if and only if the graph is planar.
• The first homology of the pure 2-braid group over a finite graph is determined by the number of disjoint cycle pairs in the graph, with explicit formulae provided in Theorem 3.25.
• For planar graphs, the 2-braid group $B_2\Gamma$ and pure 2-braid group $P_2\Gamma$ admit presentations where all relators are commutators of words corresponding to disjoint cycles.
• The second Betti number of $UD_2\Gamma$ is half that of $D_2\Gamma$, confirming $\beta_2(UD_2\Gamma) = \frac{1}{2}\beta_2(D_2\Gamma)$, which supports the commutator structure.
• The 2-braid and pure 2-braid groups over planar graphs are simple-commutator-related, as proven in Theorem 4.8, resolving a conjecture from [9].
• For $n \geq 3$, the braid group $B_n\Gamma$ is not simple-commutator-related if $\Gamma$ contains a $\Theta_4$ subgraph, leading to the conjecture that such subgraphs obstruct the property.