Reducibility of Cartesian product quantum graph equipped with group action

We consider a Cartesian product quantum graph $Γ_{n_1}\BoxΓ_{n_2}$ with standard vertex conditions, and complete the decomposition of Hilbert space $L^2(Γ_{n_1}\BoxΓ_{n_2})$ and the Laplacian $\mathscr{H}$ on it by employing the relevant theories of group representation. The concept of $Γ_{n_1}\BoxΓ_{n_2}$ equipped with the action of the cyclic group $G_{n_1}\times G_{n_2}$ is defined through the introduction of periodic quantum graph and cyclic groups. We also constructed its quotient graph and accomplish the decomposition of its secular determinant. Furthermore, under the condition that $\gcd(n_1,n_2)=1$, it can be regarded as equivalent to a circulant graph $C_{n_1n_2}(n_1,n_2)$. This work also provides a new method for the construction of isospectral graphs.

💡 Research Summary

This paper presents a systematic study on the reducibility of Cartesian product quantum graphs endowed with group actions, leveraging the power of group representation theory to decompose their associated Hilbert spaces and Laplace operators.

The work begins by establishing the necessary theoretical background. It defines a Cartesian product metric graph Γ_n1 □ Γ_n2 and rigorously formalizes what it means for a quantum graph (a metric graph paired with the Laplace operator -d²/dx² and standard vertex conditions) to be equipped with an action of a cyclic group. The action must preserve the graph structure, edge lengths, and commute with the Laplacian. The cyclic groups G_n1 and G_n2, and their direct product G_n1 × G_n2, are central to this construction.

The first major result (Theorem 4.1) addresses the decomposition of the function space. By introducing dummy vertices at the midpoints of edges to define a fundamental domain and applying representation theory, the authors prove that the Hilbert space L²(Γ_n1 □ Γ_n2) can be decomposed into a direct sum of n1 * n2 subspaces, denoted F_s,t. This decomposition is canonical with respect to the irreducible complex representations τ_s,t of the group G_n1 × G_n2. Consequently, the Laplacian H, which is self-adjoint under the standard continuity and Kirchhoff vertex conditions, can be restricted to each invariant subspace F_s,t. This effectively block-diagonalizes the spectral problem for H, breaking it down into a collection of independent, simpler problems on the quotient structure associated with each representation.

The second significant contribution involves the explicit construction of the quotient quantum graph and the factorization of its secular determinant. Using the group action, the original graph is reduced to a quotient graph whose secular determinant encapsulates the spectral information. The authors demonstrate how the secular determinant of the original Cartesian product graph factors into terms corresponding to the different irreducible representations τ_s,t. Each factor represents the secular determinant for the Laplacian restricted to the subspace F_s,t.

A particularly insightful result connects this construction to classical graph theory under a specific number-theoretic condition. When the parameters n1 and n2 are coprime (gcd(n1, n2)=1), the Cartesian product graph Γ_n1 □ Γ_n2 is shown to be isomorphic (as a graph) to a circulant graph C_{n1n2}(n1, n2). A circulant graph is one that possesses cyclic symmetry, and its spectrum is well-understood via discrete Fourier analysis. This isomorphism provides a powerful simplification: the spectral decomposition problem for the Cartesian product graph reduces to the spectral analysis of a circulant graph. This connection is not just a curiosity; it provides a novel and systematic method for constructing isospectral graphs—non-isomorphic graphs that share the same Laplace spectrum. By choosing different pairs (n1, n2) that yield the same product n1*n2 and satisfy the coprimality condition, one can obtain Cartesian product graphs that are all isomorphic to the same circulant graph and hence are isospectral to each other.

In summary, this research provides a rigorous framework for exploiting symmetry in composite quantum graph systems. It successfully applies group representation theory to decompose both the state space and the dynamics (Laplacian) of Cartesian product graphs. The link to circulant graphs under the coprimality condition offers a beautiful bridge between abstract algebraic symmetry and concrete spectral properties, yielding practical applications in the inverse spectral problem of constructing isospectral pairs.