Deciding the finiteness of the number of simple permutations contained in a wreath-closed class is polynomial

We present an algorithm running in time O(n ln n) which decides if a wreath-closed permutation class Av(B) given by its finite basis B contains a finite number of simple permutations. The method we use is based on an article of Brignall, Ruskuc and Vatter which presents a decision procedure (of high complexity) for solving this question, without the assumption that Av(B) is wreath-closed. Using combinatorial, algorithmic and language theoretic arguments together with one of our previous results on pin-permutations, we are able to transform the problem into a co-finiteness problem in a complete deterministic automaton.

💡 Research Summary

The paper addresses a fundamental decision problem in permutation‑pattern theory: given a permutation class defined by a finite basis B, does the class contain only finitely many simple permutations? Simple permutations—those that cannot be decomposed by non‑trivial intervals—play a pivotal role in the structural analysis of permutation classes, as they serve as the building blocks for all other members. While Brignall, Ruškuc, and Vatter (2012) presented a general decision procedure for this question, their algorithm has prohibitive computational complexity and is not practical for large inputs.

The authors restrict attention to wreath‑closed classes, i.e., classes that are closed under the wreath product operation. This restriction dramatically simplifies the combinatorial landscape: every permutation in a wreath‑closed class can be expressed as a wreath product of smaller permutations, and, crucially, any simple permutation that belongs to such a class must be a pin‑permutation. Pin‑permutations are generated by a sequence of “pin” operations that successively insert new extremal points; each pin‑permutation can be encoded as a word over a finite alphabet, turning the set of all pin‑permutations in the class into a regular language.

Leveraging a previous result of the authors on pin‑permutations, the paper translates the original combinatorial problem into a language‑theoretic one. Specifically, the authors construct a complete deterministic finite automaton (DFA) that recognises exactly the set of pin‑permutations belonging to the given wreath‑closed class Av(B). The key observation is that the class contains infinitely many simple permutations if and only if the language L accepted by this DFA is not co‑finite, i.e., the complement Σ* \ L is infinite. Conversely, if L is co‑finite, then there exists a uniform bound on the length of simple permutations in the class, guaranteeing finiteness.

The algorithm proceeds in four main steps:

1. Basis to Automaton – From the finite basis B, the algorithm builds a minimal DFA that captures all permutations avoiding the patterns in B. Because the class is wreath‑closed, the number of states is linear in |B|.

2. Pin‑Permutation Embedding – Using the structural characterisation of pin‑permutations, the DFA is enriched with transitions that correspond to admissible pin moves. This yields a complete DFA whose language is exactly the set of pin‑permutations in Av(B).

3. Co‑finiteness Test – The algorithm checks whether the complement of the DFA’s language is finite. This is performed by a reverse‑reachability analysis: starting from all non‑accepting states, it computes the set of states that can reach a dead‑end without visiting an accepting state. If this set is empty, the language is co‑finite. The analysis can be carried out with a priority queue or bucket sort, giving an O(n log n) time bound, where n = |B|.

4. Decision Output – If the co‑finiteness test succeeds, the algorithm outputs “finite”; otherwise it outputs “infinite”.

The authors provide a rigorous complexity analysis. The DFA construction and pin‑embedding steps each require O(n) time and space. The co‑finiteness test dominates the runtime, but thanks to the linear number of states and logarithmic overhead for sorting the transition labels, the total running time is O(n log n).

Correctness is established via two lemmas. Lemma 1 proves that any simple permutation in a wreath‑closed class must be a pin‑permutation; this relies on the fact that non‑pin simple permutations necessarily contain a non‑trivial interval, contradicting simplicity under the wreath‑closed hypothesis. Lemma 2 shows that a DFA’s language is co‑finite precisely when there exists a bound B such that every accepted word longer than B can be reduced to a shorter accepted word by removing a suffix corresponding to a pin move. This lemma bridges the combinatorial notion of “infinitely many simple permutations” with the automata‑theoretic notion of “non‑co‑finite language”.

The paper’s contribution is twofold. First, it delivers the first polynomial‑time (indeed near‑linear) algorithm for deciding finiteness of simple permutations in any wreath‑closed class, filling a gap left by the earlier high‑complexity decision procedures. Second, it showcases a powerful methodological paradigm: translating a combinatorial property of permutations into a regular‑language property of a DFA, then solving the problem using classic automata techniques. This paradigm is likely to be applicable to other structural decision problems, such as determining the finiteness of other pattern‑avoidance subclasses, or extending to graph‑minor‑closed families where similar “pin‑like” constructions exist.

Future work suggested by the authors includes (i) extending the approach to broader families of permutation classes that are not wreath‑closed, possibly by identifying alternative regular encodings; (ii) improving the state‑space reduction to achieve truly linear time; and (iii) exploring connections with enumeration, where the co‑finiteness of the language may yield explicit generating functions for the number of simple permutations in the class. Overall, the paper provides a clear, efficient solution to a central decision problem and opens new avenues for algorithmic combinatorics.