[Paper Review] Lower Bounds on the Complexity of MSO1 Model-Checking

Robert Ganian, Petr Hliněný+4 more|arXiv (Cornell University)|Sep 27, 2011

Formal Methods in Verification2 references3 citations

TL;DR

This paper establishes that MSO₁ model-checking with vertex labels is not in XP with respect to formula size for subgraph-closed graph classes whose tree-width is poly-logarithmically unbounded, unless the non-uniform Exponential Time Hypothesis (ETH) fails. The result strengthens prior lower bounds by focusing on MSO₁ instead of MSO₂, using a stronger complexity assumption to avoid reliance on effective constructions.

ABSTRACT

One of the most important algorithmic meta-theorems is a famous result by Courcelle, which states that any graph problem definable in monadic second-order logic with edge-set quantifications (i.e., MSO2 model-checking) is decidable in linear time on any class of graphs of bounded tree-width. Recently, Kreutzer and Tazari proved a corresponding complexity lower-bound - that MSO2 model-checking is not even in XP wrt. the formula size as parameter for graph classes that are subgraph-closed and whose tree-width is poly-logarithmically unbounded. Of course, this is not an unconditional result but holds modulo a certain complexity-theoretic assumption, namely, the Exponential Time Hypothesis (ETH). In this paper we present a closely related result. We show that even MSO1 model-checking with a fixed set of vertex labels, but without edge-set quantifications, is not in XP wrt. the formula size as parameter for graph classes which are subgraph-closed and whose tree-width is poly-logarithmically unbounded unless the non-uniform ETH fails. In comparison to Kreutzer and Tazari; $(1)$ we use a stronger prerequisite, namely non-uniform instead of uniform ETH, to avoid the effectiveness assumption and the construction of certain obstructions used in their proofs; and $(2)$ we assume a different set of problems to be efficiently decidable, namely MSO1-definable properties on vertex labeled graphs instead of MSO2-definable properties on unlabeled graphs. Our result has an interesting consequence in the realm of digraph width measures: Strengthening the recent result, we show that no subdigraph-monotone measure can be "algorithmically useful", unless it is within a poly-logarithmic factor of undirected tree-width.

Motivation & Objective

• To establish a complexity lower bound for MSO₁ model-checking on graph classes with unbounded tree-width.
• To extend Kreutzer and Tazari's MSO₂ lower bound to MSO₁ logic with vertex labels.
• To show that efficient MSO₁ model-checking is impossible on subgraph-closed graph classes with poly-logarithmically unbounded tree-width unless non-uniform ETH fails.
• To analyze the algorithmic limitations of directed graph width measures and their relation to undirected tree-width.
• To explore the implications for clique-width and other width parameters that are monotone under subgraphs.

Proposed method

• The proof uses a non-uniform version of the Exponential Time Hypothesis (ETH) as a complexity-theoretic assumption.
• It constructs a family of directed graphs with bounded δ-width (a hypothetical digraph width measure) but unbounded tree-width in their underlying undirected graphs.
• An advice function is used to generate digraphs that are grid-like in structure and have constant δ-width.
• The method adapts techniques from prior work to simulate MSO₁ model-checking on undirected graphs via oriented digraphs.
• It shows that if MSO₁ model-checking were in XP for such classes, then 3-Colourability could be solved in sub-exponential time with advice, contradicting non-uniform ETH.
• The argument relies on transforming MSO₁ formulas over labeled undirected graphs into equivalent formulas over oriented digraphs by replacing adjacency with symmetric arcs.

Experimental results

Research questions

• RQ1 Is MSO₁ model-checking with vertex labels in XP for subgraph-closed graph classes whose tree-width is poly-logarithmically unbounded?
• RQ2 Can a directed width measure be algorithmically useful if it largely surpasses tree-width and is subgraph-monotone?
• RQ3 What are the implications of non-uniform ETH for the tractability of MSO₁ model-checking on graph classes with unbounded tree-width?
• RQ4 Can the lower bound result be extended to unlabeled MSO₁ logic on graph classes with unbounded clique-width but efficient model-checking for bare MSO₁?
• RQ5 Is there a structural gap between tree-width and other width measures that are monotone under subgraphs?

Key findings

• MSO₁ model-checking with vertex labels is not in XP for subgraph-closed graph classes with poly-logarithmically unbounded tree-width unless non-uniform ETH fails.
• No subdigraph-monotone digraph width measure can be algorithmically useful unless it is within a poly-logarithmic factor of undirected tree-width.
• The result holds under the non-uniform ETH, which avoids the need for effectiveness assumptions used in prior proofs.
• The proof technique avoids constructing explicit obstructions by relying on advice functions and non-uniform computation.
• The findings suggest that no width parameter monotone under subgraphs can efficiently support MSO₁ model-checking on graph classes with sufficiently unbounded tree-width.
• The result implies that clique-width, despite enabling FPT model-checking for MSO₁, likely cannot be made subgraph-monotone without losing its algorithmic utility.