---
title: Circulant Quantum Markov Semigroups
url: https://www.emergentmind.com/topics/circulant-quantum-markov-semigroups
type: topic
---

# Circulant Quantum Markov Semigroups

Circulant quantum Markov semigroups (QMS) are a class of uniformly continuous quantum dynamical semigroups on finite-dimensional $C^*$-algebras exhibiting group-theoretic translation symmetry. They arise as the quantum analogue of classical circulant Markov generators, closely connected to cyclic symmetry, random walks, and the representation theory of finite abelian groups. Circulant QMS allow for explicit construction and diagonalization, making them a rare class of exactly solvable non-equilibrium quantum dynamics with fully computable quantum entropy production and detailed structural insights into irreversibility, detailed balance, and quantum currents [1210.6401].

## 1. Definitions and Algebraic Framework

Let $C_n = \{0, 1, \ldots, n-1\}$ denote the finite cyclic group of order $n$; operations are mod $n$. The function algebra $A = C(C_n) \cong \mathbb{C}^n$ and the group $C^*$-algebra $C^*(C_n)$ (generated by a unitary $U$ with $U^n = 1$) are isomorphic. In this context, a quantum Markov semigroup $(T_t)_{t \geq 0}$ on $A$ is defined as a norm-continuous family of unital, completely positive maps: $T_0 = \mathrm{Id}$ and $T_{s+t} = T_s \circ T_t$.

A QMS is called circulant (or translation-invariant) if its evolution commutes with group translation, i.e.,
$$
T_t(\delta_j) = \sum_{k \in C_n} p_t(k)\;\delta_{j+k}
$$
for some probability vector $p_t: C_n \rightarrow [0,1]$. In the group algebra picture,
$$
T_t(U^m) = \sum_k p_t(k) U^{m + k},
$$
which ensures compatibility with the coproduct $\Delta(U^m) = U^m \otimes U^m$ [1210.6768].

## 2. Lévy Processes and Generator Structure

Translation-invariant QMS on $C(C_n)$ correspond to convolution semigroups of states $(\phi_t)_{t \geq 0}$ on the associated Hopf $*$-algebra $\operatorname{Pol}(C_n) = \mathrm{span}\{U^j : 0 \leq j < n\}$, with $\Delta(U^j) = U^j \otimes U^j$, counit $\varepsilon(U^j) = 1$, and antipode $S(U^j) = U^{-j}$.

The semigroup generator is a linear functional $\ell : \operatorname{Pol}(C_n) \to \mathbb{C}$ with
$$
\ell(U^j) = -\lambda_j, \quad \ell(1) = 0, \quad \sum_j \lambda_j = 0, \quad \lambda_j \geq 0 \;\;\mbox{for}\; j \neq 0,
$$
where $\lambda_j$'s comprise the "jump rates" of the process. For a translation-invariant QMS,
$$
T_t = (id \otimes \phi_t) \circ \Delta, \quad \text{and} \quad T_t$ is circulant iff$T_t \circ \ell_h = \ell_h \circ T_t$ for all $h \in C_n$.
$$

## 3. Diagonalization and Spectral Theory

The generator $L$ of a circulant semigroup is
$$
L = \lim_{t \to 0} \frac{T_t - Id}{t}, \qquad L(\delta_j) = \sum_{r} \lambda_r \delta_{j + r}.
$$
By Fourier diagonalization, define the orthonormal Fourier basis $e_k(j) = e^{2\pi i k j / n}$, $k = 0,\ldots,n-1$, for $\mathbb{C}^n$. Then
$$
L(e_k) = -\mu_k e_k, \quad \mu_k = \sum_{r} \lambda_r (1 - e^{2\pi i k r / n}),
$$
so $L$ is diagonal in the dual basis with eigenvalues $-\mu_k$. This explicit spectrum enables tractable analysis of convergence and mixing properties [1210.6768].

## 4. Cycle Decomposition, Block-Circulant Maps, and Quantum Entropy Production

For state spaces modeled on finite abelian groups $G = \mathbb{Z}_p \times \mathbb{Z}_q$, the most general circulant completely positive (CP) map has the block-circulant structure
$$
\Phi(x) = \sum_{(i,j) \in \mathbb{Z}_p \times \mathbb{Z}_q} a(i, j)\;(J_p^i \otimes J_q^j) x (J_p^i \otimes J_q^j)^*, \quad a(i,j) \geq 0,
$$
where $J_p$ and $J_q$ are canonical left-shift permutation matrices.

The corresponding Lindblad generator is $\mathcal{L}_+(x) = \Phi(x) - x$. The semigroup's invariant state is the maximally mixed state $\rho_0 = \frac{1}{pq} I \otimes I$. The $\rho_0$-adjoint $\widetilde{\mathcal{L}}$ is given by reversing the indices: $a(p-i, q-j)$. Weighted detailed balance holds with weights $q_{ij} = a(i, j) / a(p - i, q - j)$.

The cycle decomposition represents $\Phi$ as a sum over cycles in $G$, each associated with a passage matrix, and the quantum entropy production rate (QEPR) quantifies irreversibility:
$$
e_p(\mathcal{T}, \rho_0) = \frac{1}{2pq} \sum_{(m,n) \neq (0,0)} [a(m,n) - a(p-m, q-n)] \log \frac{a(m,n)}{a(p-m,q-n)}.
$$
Quantum detailed balance holds if and only if QEPR is zero, i.e., $a(m, n) = a(p-m, q-n)$ for all $(m, n)$ [1210.6401].

## 5. Symmetry Properties: GNS, KMS, and Detailed Balance

The GNS-symmetry (self-adjointness in the GNS Hilbert space for the Haar state) is characterized by the invariance under the antipode: $\ell \circ S = \ell$, that is, $\ell(U^j) = \ell(U^{-j})$, or equivalently, $\lambda_j = \lambda_{n-j}$. On Kac-type algebras (e.g., $C_n$), KMS-symmetry with respect to the tracial Haar state coincides with GNS-symmetry.

Detailed balance (in the Alicki–Frigerio–Gorini–Verri sense) is equivalent to the symmetric condition $a(i,j) = a(p-i, q-j)$ for all $(i, j)$ and thus to the vanishing of QEPR. Weighted detailed balance interpolates between strict balance and full irreversibility [1210.6401, 1210.6768].

## 6. Potential Theory, Dirichlet Forms, and Derivation Structures

The potential theory associated with circulant QMS is encoded in the Dirichlet form
$$
\mathcal{E}(a, b) = -h(a^* L(b)),
$$
which expands explicitly to
$$
\mathcal{E}(a, b) = \frac{1}{2n} \sum_{r=1}^{n-1} \sum_{j=0}^{n-1} \lambda_r \;\overline{a(j) - a(j + r)} [b(j) - b(j + r)] \geq 0.
$$
The Schürmann triple $(\pi, \eta, \ell)$ is constructed with $\eta(U^j) = \sqrt{\lambda_j}\,\delta_j$ and $\pi$ by pointwise multiplication, leading to a derivation $\eta$ satisfying $\eta(ab) = \pi(a)\eta(b) + \eta(a)\varepsilon(b)$. The associated Dirac operator on $H \oplus L^2(C_n, h)$ realizes a possibly degenerate spectral triple, providing concrete links to noncommutative geometry [1210.6768].

## 7. Examples and Non-Equilibrium Implications

A canonical example is the $Z_p$-jump chain, $p=3$, $a(1)=\theta$, $a(2)=1-\theta$, generator
$$
\mathcal{L}_+(x) = \theta J_3 x J_3^* + (1 - \theta) J_3^* x J_3 - x,
$$
with QEPR strictly positive unless $\theta = 1/2$. For $G = \mathbb{Z}_2 \times \mathbb{Z}_2$, different choices of $a$ yield explicit deviation from detailed balance and quantified entropy production.

Circulant QMSs serve as fully explicit examples of non-equilibrium steady states (NESS) that are faithful but break detailed balance when $a(m, n) \neq a(p - m, q - n)$. The abelian symmetry allows for complete diagonalization of both the dynamics and Choi–Jamiołkowski two-point states, a feature rare in quantum dynamical semigroups [1210.6401, 1210.6768].

Circulant QMSs arise naturally in quantum information as random unitary channels invariant under cyclic shifts and in contexts requiring phase-space symmetry under the finite Weyl/Heisenberg group. These structures bridge classical cycle-based entropy production with quantum irreversibility and provide explicit laboratories for investigating NESS in open quantum systems.

Source: https://www.emergentmind.com/topics/circulant-quantum-markov-semigroups