The Cubical Homology of Trace Monoids

This article contains an overview of the results of the author in a field of algebraic topology used in computer science. The relationship between the cubical homology groups of generalized tori and homology groups of partial trace monoid actions is described. Algorithms for computing the homology groups of asynchronous systems, Petri nets, and Mazurkiewicz trace languages are shown.

💡 Research Summary

The paper develops a comprehensive homology theory for trace monoids—free partially commutative monoids—by interpreting their partial actions on sets as categorical structures and linking these to cubical (semicubical) sets and generalized tori. After introducing the basic notions of trace monoids M(E, I) defined by a set of generators E and an independence relation I, the author constructs a state space (M(E, I), S) where S is a set equipped with a partial action of the monoid. To obtain a total action, an extra “dead” state * is added, yielding the augmented state space S* = S ∪ {}. The associated small category K(S) has objects S* and morphisms given by triples (s, w, s′) with w∈M(E, I), thereby providing a categorical model for asynchronous systems and Petri nets.

The paper then introduces semicubical sets as functors from the opposite of the category □⁺ (whose objects are n‑fold products of the two‑point poset {0, 1}) to Set. Their geometric realization |X| is defined as a quotient of disjoint unions of cubes, and it is shown that for a trace monoid the “generalized torus” T(E, I) is a semicubical set whose realization is homeomorphic to the usual n‑torus when all generators are pairwise independent.

A key construction is the semicubical set Q(E, I, S) associated to a total state space: its n‑cells are pairs (x, a₁,…,aₙ) with x∈S* and (a₁,…,aₙ)∈Tₙ(E, I). The face maps either apply the corresponding generator to the current state or delete it, mirroring the partial action. The homology groups Hₙ(Q, Δℤ) (with constant integer coefficients) are shown to coincide with the homology of the state category K(S), thus reducing the computation of categorical homology to that of a semicubical set.

To compute these groups, the author employs Leech homology of the factorization category Fact(M) of a monoid M. For a trace monoid M(E, I), a functor S: □⁺/T(E, I) → Fact(M) is defined, sending each n‑tuple of independent generators to the corresponding monoid element. Theorem 3.1 proves a natural isomorphism Hₙ(Fact(M)ᵒᵖ, F) ≅ Hₙ(T(E, I), F ∘ S) for any coefficient functor F, allowing the use of a finite chain complex when E is finite. This bridges the categorical homology of the factorization category with the cubical homology of the generalized torus.

A major theoretical result concerns the vanishing of high‑dimensional homology. Let n be the maximal size of a set of pairwise independent generators in E. The paper proves that for any functor F, Hₖ(K*(S), F) = 0 for all k > n. This resolves the previously open “Problem 2” and generalizes earlier work that required finiteness of E. Consequently, the homology of asynchronous systems, Petri nets, and Mazurkiewicz trace languages is concentrated in dimensions ≤ n.

Algorithmically, the chain groups Cₙ(Q, Δℤ) are direct sums over n‑cells, and the boundary operators are expressed explicitly via the face maps. By representing these operators as integer matrices, one can compute kernels and images using standard linear algebra, thus obtaining Hₙ(K(S), Δℤ) for any desired n. The paper outlines how this procedure applies to elementary Petri nets, yielding an algorithm that computes all integral homology groups, thereby solving the open problem posed in earlier work.

In summary, the work unifies several strands—trace monoid actions, semicubical sets, generalized tori, and Leech homology—into a coherent framework that both deepens the theoretical understanding of concurrent computational models and provides concrete, implementable algorithms for their homological analysis. This bridges algebraic topology and computer science, offering tools for verification, model checking, and the study of the algebraic structure of concurrent processes.