Can You Link Up With Treewidth?

Radu Curticapean,Simon Döring,Daniel Neuen,Jiaheng Wang

2024-10-03

Abstract:A central result of Marx [ToC '10] proves that there are $k$-vertex graphs $H$ of maximum degree $3$ such that $n^{o(k /\log k)}$ time algorithms for detecting colorful $H$-subgraphs would refute the Exponential-Time Hypothesis (ETH). This result is widely used to obtain almost-tight conditional lower bounds for parameterized problems under ETH. Our first contribution is a new and fully self-contained proof of this result that further simplifies a recent work by Karthik et al. [SOSA 2024]. Towards this end, we introduce a novel graph parameter, the linkage capacity $\gamma(H)$, and show with an elementary proof that detecting colorful $H$-subgraphs in time $n^{o(\gamma(H))}$ refutes ETH. Then, we use a simple construction of communication networks credited to Beneš to obtain $k$-vertex graphs of maximum degree $3$ and linkage capacity $\Omega(k / \log k)$, avoiding the use of expander graphs. We also show that every graph $H$ of treewidth $t$ has linkage capacity $\Omega(t / \log t)$, thus recovering the stronger result of Marx [ToC '10] with a simplified proof. Additionally, we obtain new tight lower bounds for certain types of patterns by analyzing their linkage capacity. For example, we prove that almost all $k$-vertex graphs of polynomial average degree $\Omega(k^{\beta})$ for some $\beta > 0$ have linkage capacity $\Theta(k)$, which implies tight lower bounds for such patterns $H$. As an application of these results, we also obtain tight lower bounds for counting small induced subgraphs having a certain property $\Phi$, improving bounds from [Roth et al., FOCS 2020].

Data Structures and Algorithms,Computational Complexity

What problem does this paper attempt to address?

The problem that this paper attempts to solve is a key issue in complexity theory, namely, the computational complexity of detecting specific sub - graph patterns under the Exponential - Time Hypothesis (ETH). Specifically, the authors focus on how to simplify the proofs of existing results and introduce a new graph parameter - linkage capacity - to provide a new perspective for understanding these complexity issues.

### Main contributions of the paper

1. **Introduction of Linkage Capacity**:
The authors introduce a new graph parameter - the linkage capacity \(\gamma(H)\), and prove that if a multicolored \(H\)-subgraph can be detected in \(O(n^{o(\gamma(H))})\) time, then ETH will be refuted. The linkage capacity measures the ability of vertices in graph \(H\) to be connected by disjoint paths.

2. **Simplification of proofs of existing results**:
The authors provide a completely new and self - contained proof, simplifying the results of Marx [ToC ’10]. They use a simple communication network construction method (attributed to Beneš), avoiding the use of expander graphs, and thus obtain a \(k\)-vertex graph with a maximum degree of 3 and a linkage capacity of \(\Omega(k /\log k)\).

3. **New tight lower bounds**:
By analyzing the linkage capacity of certain types of graph patterns, the authors obtain new tight lower bounds. For example, they prove that for almost all \(k\)-vertex graphs with an average degree of \(\Omega(k^{\beta})\) (where \(\beta> 0\)), the linkage capacity is \(\Theta(k)\), which implies tight lower bounds for these pattern \(H\).

4. **Applications and generalizations**:
Through the linkage capacity, the authors also improve the lower bound of the time complexity for counting small induced subgraphs (with a certain property \(\Phi\)), surpassing the previous results [Roth et al., FOCS 2020].

### Formula representation

- Definition of linkage capacity: \(\gamma(H)\) is the largest constant \(c > 0\) such that for all sufficiently large \(t\in\mathbb{N}\), the graph \(H\otimes J_t\) contains a matching - connected set \(X\) and \(|X|=\lfloor ct\rfloor\).

\[
\gamma(H)=\sup\left\{c > 0\mid\forall t\in\mathbb{N},\text{ there exists }X\subseteq V(H\otimes J_t),|X| = \lfloor ct\rfloor,X\text{ is matching - connected}\right\}
\]

- Theorem 1.2: Assume that ETH holds. There exist fixed constants \(\alpha,\gamma_0 > 0\) such that for any fixed graph \(H\), if \(\gamma(H)\geq\gamma_0\), then there is no \(O(n^{\alpha\cdot\gamma(H)})\) - time algorithm to solve the \(\text{ColSub}(H)\) problem.

\[
\text{Assume ETH, there exist }\alpha,\gamma_0 > 0\text{ such that for any }H,\text{ if }\gamma(H)\geq\gamma_0,\text{ then there is no }O(n^{\alpha\cdot\gamma(H)})\text{ - time algorithm}
\]

### Summary

By introducing the new concept of linkage capacity, this paper not only simplifies the proofs of existing results but also provides new tools and methods to study the complexity issues of graph - pattern detection, especially the tightness under the ETH hypothesis.