Entropy of Quantum Markov states on Cayley trees
Extracted source text · arXiv 2208.03768
\documentclass[points]{article}
\usepackage{amsmath,amssymb}
\usepackage{tikz}
\usepackage{amssymb,amscd,amsmath,amsthm}
\usepackage{graphicx}
\usepackage{pdfsync}
\textwidth 160mm \textheight 220mm \topmargin -5mm \oddsidemargin
5mm \evensidemargin -5mm \baselineskip+6pt
%\setlength{\textheight}{9.5in} \setlength{\textwidth}{6.2in}
%setlength{\topmargin}{-.5in} \setlength{\oddsidemargin}{0in}
%\raggedbottom
%\DeclareMathOperator{\Dom}{Dom\,}
%\DeclareMathperator{\Spec}{Spec\,}
\def\bh\mathbf{h}
\def\real{\mathbf{R}}
\def\complex{\mathbf{C}}
\def\natural{\mathbf{N}}
\def\indlim{\mathop{\rm ind\,lim}}
\def\projlim{\mathop{\rm proj\,lim}}
\def\bloc{{\cal B}_{L,{\rm loc}}}
\def\ca{{\mathcal A}}
\def\cb{{\mathcal B}}
\def\cc{{\mathcal C}}
\def\cd{{\mathcal D}}
\def\ce{{\mathcal E}}
\def\cf{{\mathcal F}}
\def\cg{{\mathcal G}}
\def\ch{{\mathcal H}}
\def\ci{{\mathcal I}}
\def\cj{{\mathcal J}}
\def\ck{{\mathcal K}}
\def\cl{{\mathcal L}}
\def\cam{{\mathcal M}}
\def\cn{{\mathcal N}}
\def\co{{\mathcal O}}
\def\cp{{\mathcal P}}
\def\cq{{\mathcal Q}}
\def\car{{\mathcal R}}
\def\cs{{\mathcal S}}
\def\ct{{\mathcal T}}
\def\cu{{\mathcal U}}
\def\cx{{\mathcal X}}
\def\cz{{\mathcal Z}}
\def\cw{{\mathcal W}}
\def\bb{{\mathbb B}}
\def\bc{{\mathbb C}}
\def\bh{{\mathbf h}}
\def\bi{{\mathbb I}}
\def\bj{{\mathbb J}}
\def\bk{{\mathbb K}}
\def\bm{{\mathbb M}}
\def\bn{{\mathbb N}}
\def\bp{{\mathbb P}}
\def\bq{{\mathbb Q}}
\def\br{{\mathbb R}}
\def\bt{{\mathbb T}}
\def\bv{{\mathbb V}}
\def\by{{\mathbb Y}}
\def\bz{{\mathbb Z}}
\def\frak{\mathfrak}
\def\ga{{\frak A}}
\def\gb{{\frak B}}
\def\gc{{\frak C}}
\def\gd{{\frak D}}
\def\gf{{\frak F}}
\def\gg{{\frak G}}
\def\gh{{\frak H}}
\def\gi{{\frak I}}
\def\gj{{\frak J}}
\def\gk{{\frak K}}
\def\gl{{\frak L}}
\def\gam{{\frak M}}\def\gpm{{\frak m}}
\def\gn{{\frak N}}
\def\go{{\frak O}}
\def\gp{{\frak P}}
\def\gq{{\frak Q}}
\def\gs{{\frak S}}
\def\gt{{\frak T}}
\def\gu{{\frak U}}
\def\gv{{\frak V}}
\def\gw{{\frak W}}
\def\gx{{\frak X}}
\def\gz{{\frak Z}}
\def\a{\alpha}
\def\b{\beta}
\def\tr{{\rm Tr}}
\def\L{\Lambda}
\def\G{\Gamma}
\def\ce{\mathcal E}
\def\cA{\mathcal{A}}
\def\cH{\mathcal{H}}
\def\ffi{\varphi}
\def\cB{\mathcal{B}}
\def\bC{\mathbb{C}}
\def\bZ{\mathbb{Z}}
\def\bR{\mathbb{R}}
\def\bN{\mathbb{N}}
\def\fA{\mathfrak A}
\def\eps{\varepsilon}
\def\Tr{\mathrm{Tr}}
\def\<{\langle}
\def\>{\rangle}
\def\cH{\mathcal{H}}
\def\1{\mathbf{1}}
\def\ex{\mathrm{ex}\,}
\def\cE{\mathcal{E}}
\def\Kxy{K_{<x,y>}}
\def\Kxz{K_{<x,z>}}
\def\Hxy{H_{<x,y>}}
\def\Hxz{H_{<x,z>}}
\def\kxy{K_{<x,y>}}
\def\kxz{K_{<x,z>}}
\def\hxy{H_{<x,y>}}
\def\hxz{H_{<x,z>}}
\def\e{\varepsilon}
\def\cw{\cal W}
\def\cal{\mathcal}
\def\la{\langle}
\def\ra{\rangle}
\def\sib{\sinh \beta}
\def\cob{\cosh \beta}
\def\sibt{\sinh^2 \beta}
\def\cobte{\cosh^2 \beta}
\def\ssub{\subset \subset}
\def\subf{\subseteq_{\rm fin}}
\def\sp{\mathop{\rm sp}}
\def\spec{\mathop{\rm spec}}
\def\si{\sinh 2\beta}
\def\co{\cosh 2\beta}
%\def\sit{\sinh^2 2\beta}
%\def\cote{\cosh^2 2\beta}
%\def\s{\sigma}
%\def\de{\delta}
%\def\e{\eta}
%\def\g{\gamma}
\newcommand{\ty}[1]{\mathop{\rm {#1}}}
\def\id{{\bf 1}\!\!{\rm I}}
\def\pnorm#1#2{\left|\,#2\,\right|_{#1}}
\def\ppnorm#1#2{|\,#2\,|_{#1}}
\def\pNorm#1#2{\left\|\,#2\,\right\|_{#1}}
\def\pNORM#1#2{\left|\!\left|\!\left|\,#2\,\right|\!\right|\!\right|_{#1}}
\def\ppNorm#1#2{\|\,#2\,\|_{#1}}
\def\wexp{{\rm wexp}\,}
\def\bilin#1#2{\left\langle#1,\,#2\right\rangle}
\def\Bilin#1#2{%
\left\langle\!\left\langle#1,\,#2\right\rangle\!\right\rangle}
\def\bbilin#1#2{\langle#1,\,#2\rangle}
\def\BBilin#1#2{\langle\!\langle#1,\,#2\rangle\!\rangle}
\def\tensor#1#2{{#1}^{\otimes #2}}
\newtheorem{theorem}{Theorem}[section]
\newtheorem{proposition}[theorem]{Proposition}
\newtheorem{corollary}[theorem]{Corollary}
\newtheorem{lemma}[theorem]{Lemma}
\newtheorem{definition}[theorem]{Definition}
\newtheorem{example}[theorem]{Example}
\newtheorem{remark}[theorem]{Remark}
%\newtheorem{proof}[theorem]{Proof}
%\def\theequation{\thesection.\arabic{equation}}
%\newenvironment{proof}{\smallskip\par{\sc Proof.}\enspace}%
%{{\unskip\nobreak\hfil\penalty50\hskip2em
% \hbox{}\nobreak\hfil{\rule[-1pt]{5pt}{10pt}}
% \parfillskip=0pt\finalhyphendemerits=0
% \par\medskip}} %This is newly changed %%
%\newenvironment{remark}{\medskip\par\noindent{\bf Remark}\enspace}%
% {\medskip\par}
%%%%%%%%%%% SECTION FONT SIZE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%\makeatletter
%\def\section{\@startsection {section}{1}{\z@}{3.25ex plus 1ex minus
%.2ex}{1.5ex plus .2ex}{\large\bf}}
%\def\subsection{\@startsection{subsection}{2}{\z@}{3.25ex plus 1ex minus
% .2ex}{1.5ex plus .2ex}{\normalsize\bf}}
%\@addtoreset{equation}{section} %%
%% Without this, the numbering of equations is not reset at the beginning
%% of each section!
%%
%\makeatother
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%\document[11pt,reqno]{article}
%\documentstyle[12pt, amsfonts,amssymb,amssymb,amstex]{amsart}
\def\bloc{{\cal B}_{L,{\rm loc}}}
\def\ca{{\mathcal A}}
\def\cb{{\mathcal B}}
\def\cc{{\mathcal C}}
\def\cd{{\mathcal D}}
\def\ce{{\mathcal E}}
\def\cf{{\mathcal F}}
\def\cg{{\mathcal G}}
\def\ch{{\mathcal H}}
\def\ci{{\mathcal I}}
\def\cj{{\mathcal J}}
\def\ck{{\mathcal K}}
\def\cl{{\mathcal L}}
\def\cam{{\mathcal M}}
\def\cn{{\mathcal N}}
\def\co{{\mathcal O}}
\def\cp{{\mathcal P}}
\def\cq{{\mathcal Q}}
\def\car{{\mathcal R}}
\def\cs{{\mathcal S}}
\def\ct{{\mathcal T}}
\def\cu{{\mathcal U}}
\def\cx{{\mathcal X}}
\def\cz{{\mathcal Z}}
\def\cw{{\mathcal W}}
\def\bb{{\mathbb B}}
\def\bc{{\mathbb C}}
\def\bh{{\mathbb H}}
\def\bi{{\mathbb I}}
\def\bj{{\mathbb J}}
\def\bk{{\mathbb K}}
\def\bm{{\mathbb M}}
\def\bn{{\mathbb N}}
\def\bp{{\mathbb P}}
\def\bq{{\mathbb Q}}
\def\br{{\mathbb R}}
\def\bt{{\mathbb T}}
\def\bv{{\mathbb V}}
\def\by{{\mathbb Y}}
\def\bz{{\mathbb Z}}
\def\frak{\mathfrak}
\def\ga{{\frak A}}
\def\gb{{\frak B}}
\def\gc{{\frak C}}
\def\gd{{\frak D}}
\def\gf{{\frak F}}
\def\gg{{\frak G}}
\def\gh{{\frak H}}
\def\gi{{\frak I}}
\def\gj{{\frak J}}
\def\gk{{\frak K}}
\def\gl{{\frak L}}
\def\gam{{\frak M}}\def\gpm{{\frak m}}
\def\gn{{\frak N}}
\def\go{{\frak O}}
\def\gp{{\frak P}}
\def\gq{{\frak Q}}
\def\gs{{\frak S}}
\def\gt{{\frak T}}
\def\gu{{\frak U}}
\def\gv{{\frak V}}
\def\gw{{\frak W}}
\def\gx{{\frak X}}
\def\gz{{\frak Z}}
\def\a{\alpha}
\def\b{\beta}
\def\Tr{{\rm Tr}}
\def\L{\Lambda}
\def\G{\Gamma}
\def\ce{\mathcal E}
\def\cA{\mathcal{A}}
\def\cH{\mathcal{H}}
\def\ffi{\varphi}
\def\cB{\mathcal{B}}
\def\bC{\mathbb{C}}
\def\bZ{\mathbb{Z}}
\def\bR{\mathbb{R}}
\def\bN{\mathbb{N}}
\def\fA{\mathfrak A}
\def\eps{\varepsilon}
\def\Tr{\mathrm{Tr}}
\def\<{\langle}
\def\>{\rangle}
\def\cH{\mathcal{H}}
\def\1{\mathbf{1}}
\def\ex{\mathrm{ex}\,}
\def\cE{\mathcal{E}}
\def\Kxy{K_{<x,y>}}
\def\Kxz{K_{<x,z>}}
\def\Hxy{H_{<x,y>}}
\def\Hxz{H_{<x,z>}}
\def\kxy{K_{<x,y>}}
\def\kxz{K_{<x,z>}}
\def\hxy{H_{<x,y>}}
\def\hxz{H_{<x,z>}}
\def\ve{\varepsilon}
\def\cw{\cal W}
\def\cal{\mathcal}
\def\la{\langle}
\def\ra{\rangle}
\def\s{\sigma}
\def\bh{\mathbf{h}}
\def\bs{\mathbf{s}}
\def\bg{\mathbf{g}}
\def\e{\varepsilon}
\def\ssub{\subset \subset}
\def\subf{\subseteq_{\rm fin}}
\def\si{\sinh 2\beta}
\def\co{\cosh 2\beta}
%\def\sit{\sinh^2 2\beta}
%\def\cote{\cosh^2 2\beta}
%\def\s{\sigma}
%\def\de{\delta}
%\def\e{\eta}
%\def\g{\gamma}
%\def\Pr{\prod\!\!\!\!\!\circ}
\def\id{{\bf 1}\!\!{\rm I}}
\begin{document}
\begin{center}
{\Large {\bf Entropy of Quantum Markov states on Cayley trees}}\\[1cm]
\end{center}
\begin{center}
{\large {\sc Farrukh Mukhamedov$^*$}}\\[2mm]
\textit{ Department of Mathematical Sciences,\\
College of Science, United Arab Emirates University, \\
P.O. Box 15551,Al Ain, Abu Dhabi, UAE}\\
E-mail: {\tt far75m@yandex.ru, \ farrukh.m@uaeu.ac.ae}\\
$^*$ Corresponding author.
\end{center}
\begin{center}
{\sc Abdessatar Souissi$^*$ }\\
\textit{
Department of Accounting, College of Business Management\\
Qassim University, Buraydah, Saudi Arabia} \\
%$^2$ Preparatory institute for scientific and technical studies,\\ Carthage University, La Marsa, Tunisia}\\
E-mail: {\tt a.souaissi@qu.edu.sa}\\
$^*$ Corresponding author.
\end{center}
%\small
\begin{abstract}
In this paper, we continue the investigation of quantum Markov states (QMS) and define their mean
entropies. Such entropies are explicitly computed under certain conditions. The present work takes a huge leap forward at tackling one of the most important open problems in quantum probability which concerns the calculations of mean entropies of quantum Markov fields. Moreover, it opens new perspective for the generalization of many interesting results related to the one dimensional quantum Markov states
and chains to multi-dimensional cases.\\
\vskip 0.3cm \noindent {\it Mathematics Subject Classification}:
46L53, 60J99, 46L60, 60G50.\\
{\it Key words}: Quantum Markov state; Cayley tree; mean entropy;
\end{abstract}
\section{Introduction }\label{intr}
It is well-known that the classical and quantum dynamical entropies are important tools in many areas
of mathematics and play a huge role in information sciences \cite{OhyaPetz} and note that entropies of classical Markov chains are just as significant in statistical mechanics [10]. Quantum dynamical entropy has been studied by Connes and Sormer\cite{CS75}, Connes, Narnhofer, Thirring \cite{CNT87} and many others. Note that recently, the quantum dynamical entropy and the quantum dynamical mutual entropy were defined by Ohya in terms of complexity \cite{OhyaPetz}. The notion of quantum Markov chain (QMC) was formulated by means of the transition expectation introduced by Accardi \cite{[Ac74f]}. Therefore, in \cite{AOW97,OW19} dynamical entropy through a quantum Markov chain has been introduced and calculated for some simple models. In this similar fashion and direction a mutual entropy in quantum Markov chains scheme has been investigated in \cite{Su97}. Note the significance of these contributions as these calculations are vital in both physics and information theory.
On the other hand, the study of quantum many-body systems has lived an explosion of results. This is specifically true in the field of Tensor Networks. Recently, "Matrix Product States", and more generally
"Tensor Network States" have played a crucial role in the description of the whole quantum systems \cite{CV,Or}.
We point out that an interesting mathematical approach to quantum states on tensor networks has closely
tied up with QMC on tensor product of matrix algebras \cite{[Ac74f],GZ}. Therefore, QMC have found massive
applications in several research domains of quantum statistical physics and information \cite{fannes,F.N.W},\cite{[CiPe-Ga-Sc-Ver17]},\cite{[RoOs96]}.
As we have mentioned, it is very important to develop a method which enables to compute the entropy
for quantum systems ((\cite{CNT87,Park92}). In \cite{Park94,Park} the dynamical entropy was computed in the sense of Connes,
Narnhofer, and Thirring \cite{CNT87} of shift automorphism of QMC. Moreover, the dynamical
entropy was proved to be equal to the mean entropy. It is important to emphasize that mean entropies calculations of
QMC have been obtained in several papers \cite{Fid01,Ohn04,Petz94}. However, in all these investigates QMC were
considered over a one dimensional lattice.
On the other hand, it is crucial to know certain formulas for the mean entropies of the
Gibbs measures on each of a sequence of large finite graphs. When the graph is non-amenable, then such a calculation becomes very tricky \cite{Alp15}. Recently, in \cite{AP2018} the mentioned formula has been found for Gibbs measured defined over regular trees (in particularly, Cayley trees).
We point out that, in \cite{AOM,MBS161,MBS162,MR2004} we have started to investigate particular classes of
quantum Markov chains (QMC) associated to the Ising types models on the Cayley
trees. It turned out that the
above considered QMCs fall into a special class called
\textit{quantum Markov states (QMS)} (see \cite{MS19,MS20}). Furthermore, in \cite{FM,MS19,MS20} a description of
QMS has been carried out. It is worth to stress that the considered QMS had the
Markov property not only with respect to levels of the considered tree, but
also with regard to the interaction domain at each site, which has a finer structure, and through a family of suitable
quasi-conditional expectations which are \textit{localized}
\cite{MS19,MS21}. Such a localization property is essential for the integral decomposition of QMS, since it takes into
account the finer structure of conditional expectations and
filtration \cite{[AcFiMu07],[AcSouElG20]}.
In this paper, we continue to investigate such kind of QMS and define for these states their mean entropies and under some conditions we are able to explicitly compute them.
The present work is another step towards one of the most important open problems in quantum probability, which concerns the calculations of means entropies of quantum Markov fields.
Moreover, it opens a new perspective for the generalization of many interesting results related to one dimensional quantum Markov states and chains to multi-dimensional cases.
\section{Preliminaries}
Let $\Gamma^k_+ = (V,E)$ be a semi-infinite Cayley tree of order
$k\geq 1$ with the root $x^0$ (i.e. each vertex of $\Gamma^k_+$
has exactly $k+1$ edges, except for the root $x^0$, which has $k$
edges). Here $L$ is the set of vertices and $E$ is the set of
edges. The vertices $x$ and $y$ are called {\it nearest neighbors}
and they are denoted by $l=<x,y>$ if there exists an edge
connecting them. A collection of the pairs
$<x,x_1>,\dots,<x_{d-1},y>$ is called a {\it path} from the point
$x$ to the point $y$. The distance $d(x,y), x,y\in V$, on the
Cayley tree, is the length of the shortest path from $x$ to $y$.
Recall a coordinate structure in $\G^k_+$: every vertex $x$
(except for $x^0$) of $\G^k_+$ has coordinates $(i_1,\dots,i_n)$,
here $i_m\in\{1,\dots,k\}$, $1\leq m\leq n$ and for the vertex
$x^0$ we put $(0)$. Namely, the symbol $(0)$ constitutes level 0,
and the sites $(i_1,\dots,i_n)$ form level $n$ (i.e. $d(x^0,x)=n$)
of the lattice.
%\begin{figure}
%\begin{center}
%\includegraphics[width=10.07cm]{tree.eps}
%\end{center}
%\caption{The first levels of $\G_+^2$} \label{fig1}
%\end{figure}
The hierarchical structure on the considered tree is expressed as follows.
\[
W_n = \{ x\in V \, : \, d(x,x_0) = n\}
\qquad \Lambda_n =
\bigcup_{k=0}^n W_k, \qquad \L_{[n,m]}=\bigcup_{k=n}^mW_k, \
(n<m).
\]
For $x\in \Gamma^k_+$, $x=(i_1,\dots,i_n)$ (i.e. $x\in W_n$) define the set of its direct successors vertices by
$$
S(x) := \{y\in \Lambda_n^c:\quad <x,y> \}\subset W_{n+1}
$$
and its $k^{th}$ successors on the rooted tree $\Gamma^k_+$ is defined by the induction as follows
\begin{eqnarray*}
S_1(x) &=& S(x)\\
S_{k+1}(x) &=& S(S_k(x)),\, \, \forall k\ge 1.
\end{eqnarray*}
Put
\begin{equation}\label{S_1n_S_infty}
T(x) = \bigcup_{k\ge0}S_{k}(x)%; \quad T^{'}(x) = T(x)\setminus \{x\}.
\end{equation}
where $S_0(x) = x.$
Denote
$$
\overrightarrow{S}(x) = \{(x,1), (x,2), \cdots, (x,k)\},
$$
here $(x,i)$ means that $(i_1,\dots,i_n,i)$.
%\begin{equation}\label{binar2}
%x\circ x^0=x^0\circ x= (i_1,\dots,i_n)\circ(0)=(i_1,\dots,i_n).
%\end{equation}
%The above operation makes $\G^k_+$ a
%noncommutative semigroup with a unit. Using this semigroup
%structure one defines translations $\tau_g:\G^k_+\to \G^k_+$,
%$g\in \G^k_+$ by
%\begin{equation}\label{trans1}\tau_g(x)=g\circ x.\end{equation}
%It is clear that $\tau_{(0)}=id$.
Since the Cayley tree $\Gamma^k_+$ is regular and each vertex has exactly $k$ direct successors then according to the above enumeration, we define $k$ shifts on the tree as follows: for each $j\in \{1,2,\dots, k\}$ the $j^{th}$ shift is given by
\begin{equation}\label{shifts}
\alpha_j(x) = (j, x)= (j, i_1, i_2, \cdots, i_n)\in W_{n+1}.
\end{equation}
for every $x=(i_1, i_2, \dots, i_n)\in W_n$. Let $g = (j_1, j_2, \cdots, j_N)\in V$ one defines
$$
\alpha_g(x) := \alpha_{j_1}\circ \alpha_{j_2}\circ\cdots\circ \alpha_{j_N}(x) = (j_1, j_2, \cdots, j_N, i_1, i_2,\cdots, i_n).
$$
The shift $\alpha_g$ maps the Cayley tree $\Gamma^k_+$ onto its sub-tree $T_x$ defined by (\ref{S_1n_S_infty}).
It follows that $\alpha_x(0) = x$ and $\alpha_x(\Lambda_n)= S_n(x)$.
Let $\mathcal{A}$ be a $C^*$-algebra with unit $\id$. For each $x\in V$ we denote $\mathcal{A}_x$ the $C^*$-algebra which is the same as $\mathcal{A}$, i.e. $\mathcal{A}_x\equiv\mathcal{A}$. For each $\Lambda\subset V$ we denote by $\mathcal{A}_\Lambda$ the algebra generated by $\{\mathcal{A}_x:\, x\in\Lambda\}$. In these notations, one has
$$
\mathcal{A}_{ \Lambda_{n}} = \bigotimes_{x\in\Lambda_n}\equiv \mathcal{A}_{ \Lambda_{n}}\otimes\id_{W_{n+1}}\subset \mathcal{A}_{ \Lambda_{n+1}}.
$$
Consider the local $*-$algebra
\begin{equation}\label{AVloc}
\mathcal{A}_{V;\, loc} := \bigcup_{n\in\mathbb{N}}\mathcal{A}_{\Lambda_n}
\end{equation}
One has
$$
\mathcal{A}_V := \overline{\mathcal{A}_{V;\, loc}}^{C^*}.
$$
By $\mathcal{S}(\mathcal{B})$ we denote the set of states on a C$^*$--algebra $\mathcal{B}$.
\section{Quantum Markov states and chains on Trees}
Recall that a transition expectation is a completely positive identity preserving map from a C$^*$--algebra into its C$^*$-algebra.
\begin{definition}
Let $\mathcal{C}\subset\mathcal{B}\subset\mathcal{A}$ be a triplet of unitary C$^*$--algebras. A map $E: \mathcal{A}\to \mathcal{B}$ is called quasi-conditional expectation w.r.t. the given triplet if it is completely positive identity preserving and satisfies the following property
\begin{equation}
E(ca) = cE(a); \quad \forall a\in\mathcal{A}, \forall c\in\mathcal{C}.
\end{equation}
\end{definition}
The typical form of transitions expectations $E_{\Lambda_n}$ w.r.t. the triplet $\mathcal{A}_{\Lambda_{n-1}}\subset\mathcal{A}_{\Lambda_n}\subset\mathcal{A}_{\Lambda_{n+1}}$ that we are going to investigate in the have the following form
\begin{equation}\label{CE_TE_Ln}
{E}_{\Lambda_n} = id_{\Lambda_{n-1}}\otimes\mathcal{E}_{W_n}
\end{equation}
where
$\mathcal{E}_{W_n}$ is a transition expectation from $\mathcal{A}_{ [N,n+1] }\to \mathcal{A}_{W_n}$.
Note that $\mathcal{E}_{W_n}$ is assumed to be localized in the sense of \cite{MS19}. i.e. for each $x\in W_n$ there exists a transition expectation $\mathcal{E}_x$ from $\mathcal{A}_{\{x\}\cup S(x)}$ into $\mathcal{A}_x$ such that
\begin{equation}\label{localizedTE}
\mathcal{E}_{W_n} = \bigotimes_{x\in W_n}\mathcal{E}_x.
\end{equation}
The shifts $\alpha_j$ given by \eqref{shifts} acts on the algebra the algebra $\mathcal{A}_V$ as follows:
\begin{equation}\label{shift_algebra}
\alpha_j\left( \bigotimes_{ x\in \Lambda_{n}}a_x \right):=
\id^{(o)}\otimes\bigotimes_{i=0}^{n} a_x^{(j,x)}.
\end{equation}
\begin{definition}\label{def_TInv_MS}
A state $\varphi$ on $\mathcal{A}_{L}$ is said to be \textit{translation-invariant} if
\begin{equation}\label{trans_inv_state}
\varphi\circ \alpha_j = \varphi
\end{equation}
for every $j \in \{1,2,\cdots, k\}$.
\end{definition}
\begin{remark}
The above definition generalizes the notion of translation-invariant states on the one-dimensional lattice which corresponds to the case $k=1$. Note that the entropy of quantum Markov chains on the quantum spin lattice has been studied in many papers (see \cite{Park94, AOW97, Go98})
\end{remark}
\begin{definition}\label{QMCdef} \cite{[AcSouElG20]} Let $\phi_o\in \mathcal{S}(\mathcal{A}_o)$ be a (initial) state,
$\{E_{ \Lambda_{n}}\}$ a sequence of quasi-conditional expectations w.r.t.
the triple $\mathcal{A}_{{\Lambda}_{n-1}}\subseteq \mathcal{A}_{{\Lambda}_{n}}\subseteq\mathcal{A}_{{\Lambda}_{n+1}} $ and
a sequence $h_{n}\in\mathcal{A}_{V_n, +}$ of boundary conditions such that for each $a\in \mathcal{A}_V$ the limit
\begin{equation}\label{lim_Mc}
\varphi(a): = \lim_{n\to\infty} \phi_0\circ E_{\Lambda_{0}}\circ
E_{ \Lambda_{1}} \circ \cdots \circ E_{\Lambda_{n}}(h_{n+1}^{1/2}ah_{n+1}^{1/2})
\end{equation}
exists in the weak-*-topology and defines a state. Then the triplet $(\phi_o, (E_{ \Lambda_{n}})_n, (h_{n})_n)$ is called a quantum Markov chain on $\mathcal{A}_V$. In this case the limiting state $\varphi$ given by (\ref{lim_Mc})
is also called quantum Markov chain (QMC).
\end{definition}
\begin{definition}
A quantum Markov chain $\varphi$ on $\mathcal{A}_V$ is called \textit{quantum Markov state(QMS)} if one has
\begin{equation}\label{Markov_state_eq}
\varphi\lceil_{\mathcal{A}_{\Lambda_n}}\circ E_{\Lambda_n} = \varphi\lceil_{\mathcal{A}_{\Lambda_{n+1}}}.
\end{equation}
\end{definition}
Due to the contributions made in \cite{AFri83,GZ}, the conditional expectations can be used in place of quasi conditional expectations.
However, the boundary condition plays a more physically significant role in the existence of phase transitions (see for instance \cite{AMSa1,MBS161, MBS162, Sou21}). In \cite{MS19, MS20, MS21}, the authors studied in detail the structure of quantum Markov states associated with localized transitions expectations on the Cayley tree.
Let us now deal with the entropy of states. Assume that $\varphi$ is a state on $\mathcal{A}_V$ and $\Lambda \subset V$ is a finite region.
Let $S(\varphi\lceil_{\mathcal{A}_{\Lambda}}$) be the von Neumann entropy of the of the state $\varphi$ on the algebra $\mathcal{A}_\Lambda$ (see \cite{OhyaPetz}), where $\varphi\lceil_{\mathcal{A}_{\Lambda}}$ is the state on $\mathcal{A}_\Lambda$ defined by the restriction of $\varphi$ on $\mathcal{A}_\Lambda$. If $\mathcal{D}^{\varphi}_\Lambda$ is the density matrix of the state $\varphi\lceil\mathcal{A}_\Lambda$, then
\begin{equation}
S(\varphi\lceil_{\mathcal{A}_\Lambda}) = - \mathrm{Tr}(D_{\Lambda}^{\varphi}\log D_{\Lambda}^{\varphi})
\end{equation}
In the one dimensional case, where $V = \mathbb{N}$, using the subadditivity of $S(\varphi\lceil_{\mathcal{A}_\Lambda})$, it was shown that the limit
\begin{equation}\label{mean_entropy1D}
s(\varphi) := \lim_{n\to\infty}\frac{1}{n+1}S(\varphi\lceil_{\mathcal{A}_{[0,n]}})
\end{equation}
exists. The quantity $s(\varphi)$ is called the \textit{mean entropy} of the state $\varphi$ \cite{Fid01,Petz94}. In the next section, we are going to discuss such an entropy for quantum Markov states on the Cayley trees.
%For the sake of simplicity, the triplet $(\phi_o, \mathcal{E}, h)$ will be referred as the tree-homogeneous QMC $\varphi$.
\section{ Mean Entropy for quantum Markov states on Trees}
In this section, we introduce the mean entropy for quantum Markov states defined on trees. Let $\varphi$ be a locally faithful quantum Markov state on $\mathcal{A}_V$, i.e. $\varphi\lceil_{\mathcal{A}_{\Lambda_n}}$ is faithful for all $n\geq 1$. Indeed, by means of \cite{Ohn04} one can establish that if $\varphi$ is a locally faithful QMS, then $\varphi$ is faithful.
From the tree structure, it is quite natural to define a generalization of (\ref{mean_entropy1D}) as follows
\begin{equation}\label{mean_entropy_tree}
s(\varphi) := \lim_{n\to\infty}\frac{1}{|\Lambda_n|}S(\varphi\lceil_{\mathcal{A}_{\Lambda_n}})
\end{equation}
Let $\{E_{\Lambda_n}\}_{n\ge 0}$ be the sequence conditional expectations w.r.t. the triplet $\mathcal{A}_{\Lambda_{n-1}}\subset \mathcal{A}_{\Lambda_{n}}\subset \mathcal{A}_{\Lambda_{n+1}}$ associated with $\varphi$. We consider the Radon-Nikodym derivatives $D^{\varphi}_{\Lambda_n}$ of the state $\varphi_{\Lambda_n}:= \varphi\lceil_{\mathcal{A}_{\Lambda_n}}$ with respect to the trace $\mathrm{Tr}_{\mathcal{A}_{\Lambda_n}}$ on $\mathcal{A}_{\Lambda_n}$, i.e.
$$
\varphi_{\Lambda_n}(\cdot ) = \mathrm{Tr}_{\mathcal{A}_{\Lambda_n}}(D^{\varphi}_{\Lambda_n} \cdot )
$$
The transition expectation $\mathcal{E}_{W_n}$ associated with $E_{\Lambda_n}$ through (\ref{CE_TE_Ln}) is localized in the sense of \cite{MS19, MS20}. Then it satisfies (\ref{localizedTE}) for certain localized transition expectations $\mathcal{E}_x$ from $\mathcal{A}_{\{x\}\cup S(x)}$ into $\mathcal{A}_x$ with $x\in W_n$.
%Recall that for a given finite part $\Lambda\subset V$, denote $\mathrm{Tr_{\Lambda]}}$ the partial trace from $\mathcal{A}_{V}$ into $\mathcal{A}_{\Lambda}$ by linear extension of
%$$
%\mathrm{Tr}_{\Lambda]} (a_\Lambda\otimes a_{\Lambda'}) = \mathrm{Tr}_{\mathcal{A}_{\Lambda'}}(a_{\Lambda'}) a_{\Lambda}
%$$
%where $\Lambda'$ the complement of the subset $\Lambda$ in the full volume $V$.
For $X\subseteq Y$ two finite sub-sets of $V$, we denote $T^Y_X$ the map defined by linear extension of
$$
T^{Y}_{X}(a_X\otimes a_{Y\setminus X}) = \Tr(a_{Y\setminus X})a_X
$$
One has
$$
T^{Y}_{X}(D^{\varphi}_{Y}) = D^{\varphi}_{X}
$$
According to \cite{Mo06} the map $T_X^Y$ is the unique Umegaki conditional expectation from $\mathcal{A}_Y$ to $\mathcal{A}_X$ with respect to the tracial state.
Let $A, B$ and $C$ are mutually disjoint finite subsets of the vertex $V$. One can easily check that
\begin{eqnarray}
&&T^{A,B,C}_{A,B}\lceil_{\mathcal{A}_{B\cup C}} = T^{B,C}_{B} \label{T1}\\
&& \mathcal{A}_B = \mathcal{A}_{A\cup B}\cap \mathcal{A}_{B\cap C}\label{T2}\\
&& T^{A,B,C}_{B,C}T^{A,B,C}_{A,B} = T^{A,B,C}_{A,B}T^{A,B,C}_{B,C} = T^{A,B,C}_{C}\label{T3}
\end{eqnarray}
For each $x\in W_n$, assume that $K_{\{x\}\cup S(x)}\in\mathcal{A}_{\{x\}\cup S(x)}$ is the conditional density amplitude of the transition expectation $\mathcal{E}_x$, i.e.
$$
\mathcal{E}_x (\, \cdot \, ) = T_{\{x\}}^{\{x\}, S(x)} \left(K_{\{x\}\cup S(x)}^{*}\, \cdot\, K_{\{x\}\cup S(x)} \right)
$$
Using (\ref{localizedTE}), the conditional density amplitude of the transition expectation $\mathcal{E}_{W_n}$ has the following form
\begin{equation}\label{Knn+1=prodKxs(x)}
K_{[n,n+1]} := \bigotimes_{x\in W_n}K_{\{x\}\cup S(x)} \in\mathcal{A}_{ \Lambda_{[n,n+1]} }
\end{equation}
\begin{lemma}\label{lem_density}
Let $\varphi$ be a QMS on $\mathcal{A}_V$ together with its sequence of quasi-conditional expectations $E_{\Lambda_n}$. In the above notations, one has
\begin{equation}\label{DnDn+1}
D^\varphi_{\Lambda_{n+1}} = K_{ [n,n+1] }D^\varphi_{\Lambda_{n}} K_{ [n,n+1] }^{\ast}
\end{equation}
\end{lemma}
\begin{proof} Let $a\in \mathcal{A}_{\Lambda_{n+1}}$, one has
\begin{eqnarray*}
\varphi\lceil_{\mathcal{A}_{\Lambda_n}}(E_{\Lambda_n}(a)) &=& \mathrm{Tr}_{\Lambda_{n}}\left(D^{\varphi}_{\Lambda_n}
\mathrm{T}^{W_n, W_{n+1}}_{W_n}\left(K_{ [N,n+1] }^{\ast}a K_{ [N,n+1] }\right)\right)\\
&=& \mathrm{Tr}_{\Lambda_{n+1}}\left(D^{\varphi}_{\Lambda_n}
K_{ [N,n+1] }^{\ast}a K_{ [N,n+1] }\right)\\
&=& \mathrm{Tr}_{\Lambda_{n+1}}\left(K_{ [N,n+1] }D^{\varphi}_{\Lambda_n}
K_{ [N,n+1] }^{\ast}a \right)
\end{eqnarray*}
Using (\ref{Markov_state_eq}), we arrive at the required identity.
\end{proof}
%\section{Clustering property for quantum Markov states on trees}
%\begin{definition}
%A quantum Markov state $\varphi$ is said to satisfy the clustering property if
%\begin{equation}
%\lim_{|g|\to \infty}\varphi(a\tau_g(b)) = \varphi(a) \varphi(b)
%\end{equation}
%for all $a,b\in\mathcal{A}_{V}$.
%\end{definition}
%\begin{lemma}
%Every quantum Markov $\varphi$ associated with localized transition expectations satisfies the clustering property.
%\end{lemma}
\begin{theorem}\label{thm_entropy1} Let $\varphi$ be a QMS on $\mathcal{A}_V$. Then
\begin{equation}\label{entropy_dentity}
S(\varphi \lceil_{\mathcal{A}_{\Lambda_{n+1}}}) + S(\varphi \lceil_{\mathcal{A}_{W_{n}}}) = S(\varphi \lceil \mathcal{A}_{\Lambda_{n}})+ S(\varphi\lceil_{\mathcal{A}_{ \Lambda_{[n,n+1]} } }).
\end{equation}
If in addition $K_{ [n,n+1] }$ and $D^\varphi_{\Lambda_{n]}}$ commute, then
\begin{equation}\label{S(phi_n+1)-S(phi_n)}
S(\varphi \lceil_{\mathcal{A}_{\Lambda_{n+1}}})- S(\varphi \lceil_{\mathcal{A}_{\Lambda_{n}}}) = -2 \sum_{x\in W_{n}}\varphi_{\mathcal{A}_{ \Lambda_{[n,n+1]} }}\left( \log |K_{\{x\}\cup \overrightarrow{S}(x)}|\right).
\end{equation}
\end{theorem}
\begin{proof}
Since $\varphi$ is a Markov state on $\mathcal{A}_V$ with respect to the filtration $(\mathcal{A}_{\Lambda_n})_n$, for each $n$ there exists a quasi-conditional expectation $E_{n}$ with respect to the triplet $\mathcal{A}_{\Lambda_{n-1}}\subset\mathcal{A}_{\Lambda_n}\subset\mathcal{A}_{\Lambda_{n+1}}$ satisfying
$$
\varphi\lceil_{\mathcal{A}_{\Lambda_n}}\circ T_{\Lambda_n}^{\Lambda_{n+1}}\circ E_n = \varphi\lceil_{\mathcal{A}_{\Lambda_n}}\circ E_n = \varphi\lceil_{\mathcal{A}_{\Lambda_{n+1}}}.
$$
The quasi-conditional expectation $E_n$ can be written as follows
$$
E_n = id_{\Lambda_{n-1}}\otimes E_{[n,n+1]}
$$
where $E_{[n,n+1]}$ is a transition expectation from $\mathcal{A}_{ \Lambda_{[n,n+1]} }$ into $\mathcal{A}_{W_n}$. It follows that
$$
\varphi\lceil_{\mathcal{A}_{W_n}}\circ T_{W_n}^{W_{n+1},W_{n+1}}\circ E_{[n, n+1]} = \varphi\lceil_{\mathcal{A}_{ \Lambda_{[n,n+1]} }}\circ E_n = \varphi\lceil_{\mathcal{A}_{ \Lambda_{[n,n+1]} }}\circ E_n = \varphi\lceil_{\mathcal{A}_{ \Lambda_{[n,n+1]} }}.
$$
Then in the notations of (\ref{T1}), (\ref{T2}) and (\ref{T3}) for $A= W_{n+1}, B = W_n, C= \Lambda_{n-1}$, the conditional expectation $T^{\Lambda_{n+1}}_{\Lambda_n}$ is enough (in the sense of \cite{Petez88, Mo06}) for the states $\varphi\lceil_{\mathcal{A}_{\Lambda_{n+1}}}$ and $\varphi\lceil_{\mathcal{A}_{ \Lambda_{[n,n+1]} }}$. Hence, by a result of \cite{Mo06}, we obtain
$$
S(\varphi\lceil_{\mathcal{A}_{\Lambda_{n+1}}})
- S(\varphi_{\mathcal{A}_{W_n, W_{n+1}}}) - S(\varphi\lceil_{\mathcal{A}_{\Lambda_{n}}}) + S(\varphi\lceil_{\mathcal{A}_{W_n}}) = 0
$$
This proves (\ref{entropy_dentity}).
Now assume that $K_{ [N,n+1] }$ and $D^\varphi_{\Lambda_{n}}$ commute. Since $T^{\Lambda_{n+1}}_{W_n, W_{n+1}}( D^\varphi_{\Lambda_{n}} ) = D^\varphi_{W_n}\otimes \id_{W_{n+1}}$,
by Lemma \ref{lem_density} the density matrices of $\varphi_{\Lambda_{n+1}}$ and $\varphi_{W_n, W_{n+1}}$ are $ D^\varphi_{\Lambda_n}|K_{[n,n+1]}|^2$ and $ D^\varphi_{\Lambda_[n,n+1]} = D^\varphi_{\Lambda_n} |K_{[n,n+1]}|^2$, respectively.
This leads to
\begin{eqnarray*}
% \nonumber % Remove numbering (before each equation)
S(\varphi\lceil_{\mathcal{A}_{\Lambda_{n+1}}}) &=& -\Tr\left( D^\varphi_{\Lambda_{n}} |K_{ [N,n+1] }|^2\log ( D^\varphi_{\Lambda_{n}} |K_{ [N,n+1] }|^2)\right) \\
%&=& -\Tr\left( D^\varphi_{\Lambda_{n}} |K_{ [N,n+1] }|^2\log D^\varphi_{\Lambda_{n}}\right) -\Tr\left( D^\varphi_{\Lambda_{n}} |K_{ [N,n+1] }|^2\log |K_{ [N,n+1] }|^2\right) \\
&=& -\Tr\left( D^\varphi_{\Lambda_{n}} \log D^\varphi_{\Lambda_{n}}\mathcal{E}_{ [n,n+1] } (\id)\right)
-2\Tr\left( D^\varphi_{\Lambda_{n}}|K_{[n,n+1]}|^2 \log |K_{ [n,n+1] }|\right)\\
&=& S(\varphi\lceil_{\mathcal{A}_{\Lambda_{n}}})
-2\varphi\lceil_{\mathcal{A}_{{[n,n+1]}}}\left( \log|K_{{[n,n+1]}}| \right).
\end{eqnarray*}
On the other hand, we find
\begin{eqnarray*}
% \nonumber % Remove numbering (before each equation)
S(\varphi\lceil_{\mathcal{A}_{\Lambda_{[n, n+1]}}}) &=& - \Tr( D^\varphi_{W_n} |K_{[n,n+1]}|^2\log(D^\varphi_{W_n} |K_{[n,n+1]}|^2)) \\
%&=& - \Tr\left( D^\varphi_{\Lambda_n} |K_{[n,n+1]}|^2\left( \log\left(D^\varphi_{\Lambda_n}\right) + 2 \log\left( |K_{[n,n+1]}| \right)\right)\right) \\
% &=& - \Tr\left( D^\varphi_{\Lambda_n} |K_{[n,n+1]}|^2 \log\left(D^\varphi_{\Lambda_n}\right)\right) - 2\Tr\left(D^\varphi_{\Lambda_n} |K_{[n,n+1]}|^2 \log\left( |K_{[n,n+1]}| \right)\right) \\
&=& - \Tr\left( D^\varphi_{W_n} \log\left(D^\varphi_{W_n}\right)\mathcal{E}_{ [N,n+1] }(\id)\right) - 2\Tr\left( D^\varphi_{W_n} |K_{[n,n+1]}|^2 \log |K_{[n,n+1]}| \right) \\
&=& S(\varphi\lceil \mathcal{A}_{W_{ n}}) - 2\varphi\lceil_{\mathcal{A}_{ [N,n+1] }}\left( \log|K_{ [N,n+1] }| \right)\\
&\overset{(\ref{Knn+1=prodKxs(x)})}{=}& S(\varphi\lceil \mathcal{A}_{W_{ n}})-2 \sum_{x\in W_{n}}\varphi_{\mathcal{A}_{\Lambda_{[n, n+1]}}}\left( \log |K_{\{x\}\cup \overrightarrow{S}(x)}|\right).
\end{eqnarray*}
which proves \eqref{S(phi_n+1)-S(phi_n)}.
\end{proof}
The following result consists a generalization to trees of a formula of the mean entropy for translation-invariant quantum Markov states on the quantum spin lattice \cite{OhyaPetz}.
\begin{corollary}\label{Coro_mean} Assume that conditions of Theorem \ref{thm_entropy1} are satisfied. If a QMS $\varphi$ is translation-invariant then
\begin{equation}\label{mean_ent_formula}
\frac{S(\varphi_{\Lambda_{n+1}})- S(\varphi_{\Lambda_{n}})}{|W_n| } = S(\varphi_{\Lambda_{1]}})- S(\varphi_{\Lambda_{0]}}).
\end{equation}
\end{corollary}
\begin{proof}
One has
$$
\varphi_{\Lambda_{n+1}}(\log |K_{\{x\}\cup \overrightarrow{S}(x)}|) = \varphi_{\Lambda_{1]}}(\log |K_{\{o\}\cup \overrightarrow{S}(o)}|), \quad \forall x\in L.
$$
Then \eqref{S(phi_n+1)-S(phi_n)} leads to (\ref{mean_ent_formula}).
\end{proof}
It is known that, the mean entropy (\ref{mean_entropy1D}) of a QMS on the one-dimensional spin lattice is obtained by the following limit
$$
s(\varphi) = \lim_{n\to\infty} S(\varphi\lceil_{\mathcal{A}_{[0,n+1]}}) - S(\varphi\lceil_{\mathcal{A}_{[0,n]}})
$$
The following theorem proposes an extension of this result to the case of trees.
\begin{theorem}
The mean entropy of the Quantum Markov state $\varphi$ satisfies
\begin{equation}\label{mean_entropy_form}
s(\varphi) = \lim_{n\to \infty}\frac{S(\varphi \lceil_{\mathcal{A}_{\Lambda_{n+1}}}) -S(\varphi \lceil_{\mathcal{A}_{\Lambda_{n}}}) }{|\Lambda_{n+1}|- |\Lambda_{n}| }
\end{equation}
If in addition, $ \varphi$ is translation-invariant then
\begin{equation}\label{m_entr_hom}
s(\varphi) = \frac{S(\varphi \lceil_{\mathcal{A}_{\Lambda_{1}}}) -S(\varphi \lceil_{\mathcal{A}_{\Lambda_{0}}})}{k}.
\end{equation}
\end{theorem}
\begin{proof}
One has %$|\Lambda_n| = |\Lambda_{n+1}| - (k-1)|\Lambda_{n}|$ then
\begin{eqnarray*}
\frac{S(\varphi \lceil \mathcal{A}_{\Lambda_{n+1}}) -S(\varphi \lceil \mathcal{A}_{\Lambda_{n}}) }{|\Lambda_{n+1}| - |\Lambda_{n}|}
&=& \frac{|\Lambda_{n+1}|}{|\Lambda_{n+1}|}\frac{S(\varphi \lceil \mathcal{A}_{\Lambda_{n+1}})}{|\Lambda_{n+1}|}-\frac{|\Lambda_{n}|}{|\Lambda_{n+1}|}
\frac{S(\varphi \lceil \mathcal{A}_{\Lambda_{n}}) }
{|\Lambda_{n}|} \\
&=& \frac{k^{n+2}-1}{(k-1)k^{n+1}}\frac{S(\varphi \lceil \mathcal{A}_{\Lambda_{n+1}})}{|\Lambda_{n+1}|}-\frac{k^{n+1}-1}{(k-1)k^{n}}
\frac{S(\varphi \lceil \mathcal{A}_{\Lambda_{n}})}
{|\Lambda_{n}|}.
\end{eqnarray*}
This leads to (\ref{mean_entropy_form}).
In the translation-invariant case, by Corollary \ref{Coro_mean} we arrive at (\ref{m_entr_hom}).
\end{proof}
\begin{example} Let us consider the semi-infinite Cayley tree of order two $\Gamma^+_2$. Let $\mathcal{A}_{x}=M_{2}(\mathbb{C})$ for $x\in \mathcal{A}_{V}$. By $\sigma_{x}^{u}$,
$\sigma_{y}^{u}$, $\sigma_{z}^{u}$ we denote the standard Pauli spin matrices at site $u\in V$, i.e.
$$
\id^{(u)}=\begin{pmatrix}
1 & 0\\
0& 1
\end{pmatrix},\ \sigma_{x}^{(u)}=\begin{pmatrix}
0 & 1\\
1& 0
\end{pmatrix} , \ \sigma_{y}^{(u)}=\begin{pmatrix}
0 & -i\\
i& 0
\end{pmatrix}, \ \sigma_{z}^{(u)}=\begin{pmatrix}
1 & 0\\
0& -1
\end{pmatrix}.
$$
For every vertices $(u,(u,1),(u,2))$ we put
\begin{eqnarray}\label{1Kxy1}
&&K_{<u,(u,i)>}=\exp\{\b H_{u,(u,i)>}\}, \ \ i=1,2,\ \b>0,\\[2mm] \label{1Lxy1} &&
L_{>(u,1),(u,2)<}=\exp\{J\beta H_{>(u,1),(u,2)<}\}, \ \ J>0,
\end{eqnarray}
where
\begin{eqnarray}\label{1Hxy1}
&&
H_{<u,(u,i)>}=\frac{1}{2}\big(\id^{(u)}\id^{(u,i)}+\sigma^{(u)}_z\sigma^{(u,i)}_z\big),\\[2mm]
\label{1H>xy<1} &&
H_{>(u,1),(u,2)<}=\frac{1}{2}\big(\id^{(u,1)}\id^{(u,2)}+\sigma^{(u,1)}_z\sigma^{(u,2)}_z\big).
\end{eqnarray}
For each $n\in \mathbb{N}$ and $x\in V$, we denote
\begin{eqnarray}\label{K1}
&& A_{x,(x,1),(x,2)}=K_{<x,(x,1)>}K_{<x,(x,1)>}L_{>(x,1),(x,2)<},\\[2mm]
\label{K11} &&K_{[m,m+1]}:=\prod_{x\in \overrightarrow
W_{m}}A_{x,(x,1),(x,2)}, \ \ 1\leq m\leq n,\\[2mm]
\label{K2} &&
\bh_{n}^{1/2}:=\prod_{x\in\overrightarrow
W_n}(h_{x})^{1/2}, \ \ \ \bh_n=\bh_{n}^{1/2}(\bh_{n}^{1/2})^*\\[2mm]
\label{K3}
&&{\mathbf{K}}_n:=\omega_{0}^{1/2}\prod_{m=1}^{n-1}K_{[m,m+1]}\bh_{n}^{1/2}\\[2mm]
\label{K4} && \mathcal{W}_{n]}:={\mathbf{K}}_n^{*}{\mathbf{K}}_n.
\end{eqnarray}
Here $w_0\in\mathcal{A}_{o, +}$ is and initial state and $h_x\in\mathcal{A}_{x,+}$ is a family of positive boundary conditions.
Consider the following functional
$\ffi^{(n)}_{w_0,\bh}$, given by
\begin{eqnarray}\label{ffi-ff}
\varphi^{(n)}_{w_0,\bh}(a)=\tr(\cw_{n+1]}(a\otimes\id_{W_{n+1}})),
\end{eqnarray}
for every $a\in \cb_{\Lambda_n}$.
In \cite{MBS161} we have shown that the compatibility equation (\ref{Markov_state_eq}) is satisfied for the sequence $\varphi^{(n)}_{w_0, \bh}$ under the following conditions:
\begin{eqnarray}\label{eq1}
&&\Tr(\omega_{0}h_{0})=1, \\
\label{eq2} &&\Tr_{x]}\big({A_{x,(x,1),(x,2)}}h_{(x,1)}h_{(x,2)}
A_{x,(x,1),(x,2)}^{*}\big)=h_x, \ \ \textrm{for every}\ \ x\in L,
\end{eqnarray}
Moreover, it has been established the existence of three quantum Markov states $\varphi_1, \varphi_2$ and $\varphi_\alpha$ associated with three different translation-invariant solutions $h_1, h_2$ and $h_\alpha$ of (\ref{eq1}) with
$$
\alpha = \frac{4}{e^{2J\beta}(e^{4\beta}+1)+2e^{2\beta}}.
$$
In the following we will compute the mean entropy for the Markov state $\varphi_\alpha$. In \cite{MS201} we found the density matrices $\mathcal{W}_{n]}^{(\alpha)}$ of $\varphi_\alpha$ as follows
$$\mathcal{W}_{n]}^{(\varphi)} = \exp\mathcal{H}_{n]}^{(\alpha)}$$
\begin{equation}\label{densityphialpha}
\mathcal{H}_{n]}^{(\alpha)}:=\sum_{x\in
W_{n-1]}}\mathcal{H}_{\alpha}^{(x,(x,1),(x,2))}
\end{equation}
where
$$\mathcal{H}_{\alpha}^{(x,(x,1),(x,2))}:=\id^{\otimes |\{<_{(W_{n-1})}x\}|}\otimes\mathcal{H}^{(x)}_{\alpha}\otimes\id^{\otimes(|\{>_{(W_{n-1})}x\}|-2)},$$
here \small
$$\mathcal{H}_{\alpha}:= \ln\a\id_{M_8(\mathbb{C})}+
2\left(
\begin{array}{cccccccc}
\beta(J+2) & & & & & & & \\
& \beta & & & & & & \\
& & \beta & & & \left(0\right) & & \\
& & & \beta J & & & & \\
& & & & \beta J & & & \\
& & \left(0\right) & & & \beta & & \\
& & & & & & \beta & \\
& & & & & & & \beta(J+2) \\
\end{array}
\right)$$
\normalsize and $\{<_{(W_{n-1})}x\}:=\{y\in W_{n-1}\mid y<x\}$,
$\{>_{(W_{n-1})}x\}:=\{y\in W_{n-1}\mid y>x\}$. In this notation,
the order $"x<y"$ is understood with respect to the lexicographic
order applied to the coordinates of $x$ and $y$.
It follows that
\begin{equation}\label{densityn}
\mathcal{W}_{n]}^{(\varphi_\alpha)} = \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\exp\left(\sum_{x\in W_{n-1}}\mathcal{H}_{\alpha}^{(x,(x,1),(x,2))}\right)
\end{equation}
The von Neuman entropy of $\varphi^{\alpha}\lceil_{\mathcal{A}_{\Lambda_n}}$ satisifies
\begin{eqnarray*}
S(\varphi_{\alpha}\lceil_{\mathcal{A}_{\Lambda_n}})
&= & -\tr\left( \mathcal{W}_{n]}^{(\varphi_\alpha)}\log \mathcal{W}_{n]}^{(\varphi_\alpha)}\right)\\
&\overset{(\ref{densityn})}{=}& -\tr\left( \mathcal{W}_{n]}^{(\varphi_\alpha)}\left(\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)} + \sum_{x\in W_{n-1}}\mathcal{H}_{\alpha}^{(x,(x,1),(x,2))} \right)\right) \\
&=& -\tr\left( \mathcal{W}_{n]}^{(\varphi_\alpha)}\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\right) -\sum_{x\in W_{n-1}}\tr\left( \mathcal{W}_{n]}^{(\varphi_\alpha)} \mathcal{H}_{\alpha}^{(x,(x,1),(x,2))} \right)\\
&=& -\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_n}}\left(\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\right) -\sum_{x\in W_{n-1}}\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_n}}\left( \mathcal{H}_{\alpha}^{(x,(x,1),(x,2))} \right) \\
&=& -\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_{n-1}}}\left(\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\right) -\sum_{x\in W_{n-1}}\varphi_\alpha\lceil_{\mathcal{A}_{\{x\}\cup S(x)}}\left( \mathcal{H}_{\alpha}^{(x,(x,1),(x,2))} \right)
\end{eqnarray*}
One has
$$
- \varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_{n-1}}}\left(\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\right) = -\tr\left( \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\log \mathcal{W}_{n-1]}^{(\varphi_\alpha)}\right) = S(\varphi_{\alpha}\lceil_{\mathcal{A}_{\Lambda_{n-1}}})
$$
and since the state $\varphi_\alpha$ is translation-invariant then
$$
\varphi_\alpha\lceil_{\mathcal{A}_{\{x\}\cup S(x)}}\left( \mathcal{H}_{\alpha}^{(x,(x,1),(x,2))} \right) = \varphi_\alpha\lceil_{\mathcal{A}_{\{x_0\}\cup S(x_0)}}\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))} \right)
$$
Therefore
$$
S(\varphi_{\alpha}\lceil_{\mathcal{A}_{\Lambda_{n}}}) =
S(\varphi_{\alpha}\lceil_{\mathcal{A}_{\Lambda_{n-1}}}) - |W_{n-1}|\varphi_\alpha\lceil_{\mathcal{A}_{\{x_0\}\cup S(x_0)}}\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))} \right).
$$
This leads to (\ref{mean_ent_formula}). Using (\ref{m_entr_hom}) the mean entropy of the QMS $\varphi_\alpha$ is given by
\begin{eqnarray*}
s(\varphi_\alpha) &=& \frac{1}{2}\varphi_\alpha\lceil_{\mathcal{A}_{\{x_0\}\cup S(x_0)}}\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))} \right)\\
&=& \frac{1}{2} \tr\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))} \exp\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))}\right)\right)\\
&=& \alpha\left( (\ln \alpha + 2\beta(J+2))e^{2\beta (J+2)} + 2 (\ln \alpha + 2\beta)e^{2\beta} + (\ln \alpha + 2\beta J)e^{2\beta J}\right)
\end{eqnarray*}
We can compute the mean entropy of QMS $\varphi_\alpha$ directly from (\ref{mean_entropy_tree}). In fact,
\begin{eqnarray*}
S(\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_n}}) &=& - \tr\left(\mathcal{W}_{n]}^{(\varphi_\alpha)}\log \mathcal{W}_{n]}^{(\varphi_\alpha)} \right)\\
& \overset{(\ref{densityphialpha})}{=}& -
\sum_{x\in \Lambda_{n-1}}
\tr\left(\mathcal{W}_{n]}^{(\varphi_\alpha)} \mathcal{H}_{\alpha}^{(x ,(x ,1),(x ,2))}\right)\\
&=& -
\sum_{x\in \Lambda_{n-1}}
\varphi_\alpha\lceil_{\mathcal{A}_{\{x\}\cup S(x)}}\left( \mathcal{H}_{\alpha}^{(x ,(x,1),(x,2))}\right)\\
&=& -
|\Lambda_{n-1}|
\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_1}}\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))}\right)
\end{eqnarray*}
Then
\begin{eqnarray*}
s(\varphi_\alpha) &=& \lim_{n\to \infty}\frac{S(\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_n}})}
{|\Lambda_n|} = \lim_{n\to \infty}\frac{S(\varphi_\alpha\lceil_{\mathcal{A}_{\Lambda_n}})}
{|\Lambda_{n-1}|}\frac{|\Lambda_{n-1}|}{|\Lambda_n|} \\
&=& -\frac{1}{2}\tr\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))} \exp\left( \mathcal{H}_{\alpha}^{(x_0,(x_0,1),(x_0,2))}\right)\right).
\end{eqnarray*}
\end{example}
\section{Mixing property for quantum Markov states on Cayley trees}
In this section, we consider the semi-infinite Cayley tree $\Gamma^k_+$ and, it is assumed that $\mathcal{A}_x = M_d$ the $d\times d$ square matrices over the complex field $\mathbb{C}$. Let $\varphi$ be a QMS on $\mathcal{A}_V$ together with its sequence of localized transition expectations $\mathcal{E}_{W_n} = \bigotimes_{x\in W_n}\mathcal{E}_x$, here $\mathcal{E}_x$ is a transition expectation from $\mathcal{A}_{\{x\}\cup S(x)}$ into $\mathcal{A}_x$. Put $\mathcal{B}_x = \mathcal{E}_x(\mathcal{A}_{\{x\}\cup S(x)})\subseteq\mathcal{A}_x$.
If the QMS $\varphi$ is translation-invariant, then the transition expectations $\mathcal{E}_x$ can be considered as copies of a transition expectation $\mathcal{E}$ from $M_d^{\otimes (k+1)}$ into $M_d$. In this setting, we denote $\mathcal{B}= \mathcal{E}(M_d^{\otimes (k+1)})$.
Now, let $p_1,\dots,p_r$ be the minimal central projections (counted with their multiplicities) of $\mathcal{B}$, so that the reduced algebras $p_i\mathcal{B}p_i\cong M_{d_i}$ for some $d_i\in\mathbb{N}$. Then
$$
\mathcal{B} = \bigoplus_{i=1}^{r}p_i\mathcal{B}p_i= \bigoplus_{i=1}^{r}M_{d_i}.
$$
Define
$$
E(A) = \sum_{i=1}^{r} p_iAp_i
$$
Put $\mathbb{T} = \{z\in\mathbb C:\quad |z|=1 \}.$
%Let $\varphi$ be a quantum Markov state on $\mathcal{A}_V$ together with its sequence of localized transition expectations (\ref{localizedTE}) \\
\begin{definition}
A QMS $\varphi$ on $\mathcal{A}$ is said to be \textit{strongly mixing} if it satisfies
$$
\lim_{|g|\to\infty} \varphi(a \alpha_g(b)) = \varphi(a)\varphi(b)
$$
for all $a,b\in\mathcal{A}_V$.
\end{definition}
Any translation-invariant QMS $\varphi$ on $\mathcal{A}_V$ satisfies
\begin{equation}
\varphi\lceil_{\mathcal{A}_{V_g}}\circ\, \alpha_g = \varphi
\end{equation}
where $V_g = \alpha_g(V)$.
\begin{theorem}\label{thm_mixing}
Every faithful, translation invariant QMS with localized transition expectations on $\mathcal{A}_V$ is strongly mixing.
\end{theorem}
\begin{proof}
Let $\varphi$ be a translation-invariant QMS on $\mathcal{A}_V$ together with its sequence of localized transition expectations $\mathcal{E}_{W_n}=\bigotimes_{x\in W_n}\mathcal{E}_{x}$ as in (\ref{localizedTE}), here $\mathcal{E}_x\equiv \mathcal{E}$ is a Umegaki conditional expectation from $\mathcal{A}_{\{x\}\cup S(x)}$ into $\mathcal{A}_x$. For each $x\in W_n$ and $j\in\{1,2,\dots, k\}$, we put
$$
P_{j}= P_{x; j}:= \mathcal{E}_x\lceil_{\mathcal{A}_x\otimes \mathcal{B}^{(x,j)}\otimes\id_{S(x)\setminus \{(x,j)\}}}.
$$
One can see that $P_{x; j}$ is completely positive and identity preserving from $\mathcal{B}$ into itself. Let $\lambda\in\mathbb{T}$ and $B\in \mathcal{B}\setminus \{0\}$ such that $P_{x;j}(B^{(x,j)})= \lambda B^{(x)}$. Then for every $A' \in \mathcal{B}$,
\begin{eqnarray*}
\lambda (B'B)^{(x)} &=& (B'^{(x)}\otimes \id_{S(x)}) P_{x;j}(B^{(x,j)})\\
&=& (B'^{(x)}\otimes \id_{S(x)}) \mathcal{E}_x(\id\otimes B^{(x,j)})\\ &=& \mathcal{E}_x(B'^{(x)}\otimes\id_{S(x)\setminus \{(x,j)\}}\otimes B^{(x,j)})\\
&=& \mathcal{E}_x( \id_{\{x\}\cup S(x)\setminus }\otimes B^{(x,j)})B'^{(x)}\otimes\id_{S(x)} \\
&=& \lambda (BB')^{(x)}\otimes \id_{S(x)}.
\end{eqnarray*}
Then $B$ belongs to the center of $\mathcal{B}$, and there exist $\beta_i\in\mathbb{C} (1\le i\le r$) such that $B = \sum_{i=1}^{r}\beta_ip_i$. From \cite{AcLib99, MS19, MS20} there exist positive linear functional $\rho_{ii'}$ on $M_i\otimes M_{i'}$ such that
\begin{equation}\label{rhoii'}
\mathcal{E}_x(p_i^{(x)}\otimes p_{i'}^{(x,j)}\otimes \id_{S(x)\setminus\{(x,j)\}}) = \rho_{ii'}(p_i\otimes p_{i'}) p_{i}
\end{equation}
One has
\begin{eqnarray*}
P_{x,j}(B) &=& \mathcal{E}_x(\id^{(x)}\otimes B^{(x,j)}\otimes\id_{S(x)\setminus\{(x,j)\}})\\
&=& \sum_{ii'}\mathcal{E}_x(p_{i}\otimes \beta_{i'}p_{i'} \otimes \id_{S(x)\setminus\{(x,j)\}})\\
&=& \sum_{ii'}\beta_{i'}\mathcal{E}_x(p_{i}\otimes p_{i'} \otimes \id_{S(x)\setminus\{(x,j)\}})\\
&\overset{(\ref{rhoii'})}{=}& \sum_{ii'}\beta_{i'}\pi_{ii'}p_{i}\\
\end{eqnarray*}
where $\beta_{ii'} = \rho_{ii'}(p_i\otimes p_{i'})>0$. We have
$$
\left[ \begin{array}{ccc}
\pi_{11} & \cdots & \pi_{1r}\\
\vdots & \ddots & \vdots\\
\pi_{r1} & \cdots & \pi_{rr}
\end{array}\right]
\left[ \begin{array}{lll}
\beta_{1} \\
\vdots \\
\beta_{r}
\end{array}\right] =
\lambda \left[ \begin{array}{lll}
\beta_{1} \\
\vdots \\
\beta_{r}
\end{array}\right]
$$
Since $\pi_{ii'}>0$ for all $ii'$ then by the Perron-Frobenius theorem the map $P_{x,j}$ has trivial peripheral spectrum. Therefore, by \cite{F.N.W} the QMS $\varphi$ is strongly mixing.
\end{proof}
Now, we are going to establish an analogue of \cite[Theorem 3.1]{Go98}.
\begin{theorem}\label{main}
Let $\varphi$ be a locally faithful QMS on $\mathcal{A}_V$. Let $(\mathcal{H}, \pi, \Omega )$ be its associated GNS triplet.
The following assertions hold true.
\begin{itemize}
\item[(i)] The quantum Markov state $\varphi$ on the quasi-local algebra $\mathcal{A}_V$ is separating.
i.e. the cyclic vector $\Omega$ is cyclic for the algebra $\mathcal{B}:= \pi(\mathcal{A}_V)^{''}$.
\item[(ii)] The algebra $\mathcal{B} =\pi(\mathcal{A}_V)^{''} $ is a factor.
% \item[(iii)] The centralizer $\mathcal{N}_{\varphi}$ is a hyperfinite type II factor.
\end{itemize}
\end{theorem}
\begin{proof} (i).
Let $(\mathcal{H}, \pi, \Omega)$ be the GNS-triplet associated with the state locally faithful state $\varphi$.
Let $a\in\mathcal{A}_{\Lambda_{n}}$, we define an operator $\hat{a}$ as
$$
\hat{a}\pi(b)\Omega = \pi(ba)\Omega; \quad \forall b\in \mathcal{A}_{L, loc}.
$$
The operator $\hat{a}$ acts on $\pi(\mathcal{A}_{L})$ and commutes with $\pi(b)$ for every $b\in \mathcal{A}_{L, loc}$.
On the other hand, we have
\begin{eqnarray*}
\langle \hat{a} \pi(b) \Omega, \hat{a} \pi(b) \Omega \rangle
%& =& \langle \pi(ba) \Omega, \pi(ba) \Omega \rangle \\
&=&\varphi(a^\ast b^\ast ba ) \\
&=& \varphi(E_{n+1]}(a^\ast b^\ast ba))\\
&=& \varphi(a^\ast E_{n+1]}( b^\ast b)a)\\
&=& \Tr\left((D_{\Lambda_{n+1}}a^\ast E_{n+1]}( b^\ast b)a \right)\\
&=& \Tr\left(D_{\Lambda_{n+1}}^{\varphi})^{1/2}
E_{n+1]}( b^\ast b)(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\Big|(D_{\Lambda_{n+1}}^{\varphi})^{-1/2}a(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\Big|^2\right)\\
&\le & \Tr\left((D_{\Lambda_{n+1}}^{\varphi})^{1/2}
E_{n+1]}( b^\ast b)(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\right)\Big\|(D_{\Lambda_{n+1}}^{\varphi})^{-1/2}a(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\Big\|^2.
\end{eqnarray*}
and thanks to (\ref{Markov_state_eq}), one get
\begin{eqnarray*}
\Tr\left((D_{\Lambda_{n+1}}^{\varphi})^{1/2}
E_{n+1]}( b^\ast b)(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\right)& =& \Tr\left( D_{\Lambda_{n+1}}^{\varphi}
E_{n+1]}( b^\ast b)\right)\\
&=& \varphi( E_{n+1]}( b^\ast b))\\
&=& \varphi(b^\ast b)\\
&=& \| \pi(b) \Omega\|^2.
\end{eqnarray*}
Thus
$$
\|\hat{a}\| \le \Big\|(D_{\Lambda_{n+1}}^{\varphi})^{-1/2}a(D_{\Lambda_{n+1}}^{\varphi})^{1/2}\Big\|.
$$
It follows that $\hat{a}\in \pi(\mathcal{A}_L)'$. The set $\{\hat{a}\Omega= \pi(a) \Omega : a\in \mathcal{A}_{L, loc} \}$ is dense in $\mathcal{H}$. This means that
$\Omega$ is cyclic for $\pi(\mathcal{A}_L)^{'}$, equivalently $\Omega$ is separating for $\pi(\mathcal{A}_L)^{''}$. This proves (i).
Now according to Theorem \ref{thm_mixing}, from \cite{BR} we arrive at (ii).
%(ii). Since for each $x\in\Lambda_n$ the quasi-conditional expectation $E_x$ w.r.t. the triplet
% $\mathcal{A}_{P(x)}\subset \mathcal{A}_{\{x\}\cup P(x)}\subset \mathcal{A}_{P(x)\cup \{x\}\cup S(x)}$ preserves the state $\varphi_x:= \varphi_{\lceil \mathcal{A}_{P(x)\cup \{x\}\cup S(x)}}$, there exists a $\sigma_t^{\varphi_x}$-invariant
% C$^*$--subalgebra $\mathcal{N}_x$ such that $\mathcal{A}_{\{x\}\cup P(x)}\subseteq \mathcal{N}_x \subseteq\mathcal{A}_{P(x)\cup \{x\}\cup S(x)}$.
%
% Since the algebra $\mathcal{A}_{P(x)\cup \{x\}}$ is a factor of type I and $\mathcal{A}_{S(x)}$ is a finite von Neumann algebra,
% in accordance with \cite{LiG95} there exists a von Neumann subalgebra $\widetilde{\mathcal{N}}_{x}$ of $\mathcal{A}_{S(x)}$ such that
% $\mathcal{N}_{x} = \mathcal{A}_{\{x\}\cup P(x)}\otimes\widetilde{\mathcal{N}}_{x}.$\\
% There exists a $\varphi_x$-preserving conditional expectation $F_x$ from $\mathcal{A}_{P(x)\cup T(x)}$ onto $\mathcal{N}_{x}$.
% Let $\widetilde{\mathcal{N}}_{0,x} = \alpha_x^{-1}(\widetilde{\mathcal{N}}_{x})$.
% The map $F_x$ is then valued $\mathcal{A}_{\{x\}\cup P(x)}\otimes \alpha_x(\widetilde{\mathcal{N}}_{0,x})$.
% Let $N = \bigcup_{x\in V}\bigcap_{y\in T_x}\mathcal{N}_{0,y}$. There exists a conditional expectation
%
% The algebra $\mathcal{N}_{\Lambda_{ n+1]}}$ is then generated by $\mathcal{A}_{\Lambda_{n}}$ and its relative
% commutant $\mathcal{A}^{'}_{\Lambda_{n}}\cap \mathcal{N}_{\Lambda_{n+1} }= \mathcal{A}_{\Lambda_{n+1}}\cap \mathcal{N}_{\Lambda_{n+1} }$.
\end{proof}
%\section{Conclusion}
\section*{Credit Author Statement}
The all authors of this paper are equally contributed for the realization of the results.
\section*{Declaration of Competing Interest}
The authors confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.
\section*{Data availability}
The paper does not use any data.
\section*{Acknowledgments}
The authors gratefully acknowledge Qassim University, represented by the Deanship of Scientific Research,
on the financial support for this research under the number (10173-cba-2020-1-3-I) during the academic year 1442 AH / 2020 AD.
\begin{thebibliography}{49}
\bibitem{[Ac74f]}
Accardi L., On the noncommutative Markov property, \emph{Funct.
Anal. Appl.}, {\bf 9} (1975) 1--8.
\bibitem{[AcFiMu07]}
Accardi L., Fidaleo F., Mukhamedov, F., Markov states and chains on
the CAR algebra, \textit{Inf. Dim. Analysis, Quantum Probab. Related
Topics} {\bf 10} (2007), 165--183.
\bibitem{AFri83} Accardi, L., Frigerio, A., Markovian cocycles, \textit{Proc. R. Irish Acad. A} {\bf 83}(1983), 251–263 (1983).
\bibitem{AcLib99} Accardi, L., Liebscher, V., Markovian KMS-states for one-dimensional spin chains, Infin. Dimens. Anal. Quantum
Probab. Relat. Top., 2: 645–661 (1999)
\bibitem{AMSa1} Accardi L., Mukhamedov, F. Saburov M. On Quantum Markov Chains on Cayley tree I:
uniqueness of the associated chain with $XY$-model on the Cayley tree of order two,
\textit{Inf. Dim. Analysis, Quantum Probab. Related Topics} {\bf 14}(2011), 443--463.
\bibitem{AOM} Accardi L., Ohno, H., Mukhamedov F., Quantum Markov fields on
graphs, \textit{Inf. Dim. Analysis, Quantum Probab. Related Topics}
{\bf 13}(2010), 165-189.
\bibitem{AOS94} Accardi L., Ohya M., Suyari H., Computation of Mutual Entropy in Quantum
Markov Chains, \textit{Open Sys. \& Infor. Dyn.} {\bf 2} (1994),337-354.
\bibitem{AOW97} Accardi L., Ohya M., Watanabe N., Dynamical entropy through quantum Markov chains, \textit{Open Sys.
\& Infor. Dyn.} {\bf 4} (1997), 71-87.
\bibitem{[AcSouElG20]} Accardi L., Souissi A., Soueidy E. G. Quantum Markov chains: A unification approach, \textit{Inf. Dim. Analysis, Quantum Probab. Related Topics } 23(2020).
.
%\bibitem{AF94} Alicki, R., Fannes, M. Defining quantum dynamical entropy, \textit{Lett. Math.
%Phys.} {\bf 32} (1994), 75–82.
\bibitem{Alp15} Alpeev, A., The entropy of Gibbs measures on sofic groups, \textit{Zap. Nauchn. Sem.
POMI}, {\bf 436}(2015), 34–48.
\bibitem{AP2018} Austin T., Podder M. Gibbs Measures Over Locally Tree-Like Graphs and
Percolative Entropy Over Infinite Regular Trees, \textit{J. Stat. Phys.} {\bf 170} (2018), 932–951.
\bibitem{Bax} Baxter R. J. \textit{Exactly Solved Models in Statistical Mechanics},
(Academic Press, London-New York 1982)
\bibitem{BR} Bratteli O., Robinson D.W., \textit{Operator algebras and quantum statistical
mechanics I}, Springer-Verlag, New York, 1987.
\bibitem{BR2} Bratteli O., Robinson D.W., \textit{Operator algebras and quantum statistical
mechanics II}, Springer-Verlag, New York, 1987.
\bibitem{[CiPe-Ga-Sc-Ver17]}
Cirac J. I., Perez-Garcia D., Schuch N., Verstraete F.,
Matrix Product Unitaries: Structure, Symmetries, and Topological Invariants,
\textit{J. Statis. Mech. Theory and Experiment} (2017), 083105.
\bibitem{CV} Cirac J.I., Verstraete F. Renormalization and tensor product states in spin chains and
lattices, \textit{J. Phys. A. Math. Theor.} {\bf 42} (2009),
504004.
\bibitem{CS75} Connes A., Stormer E., Entopy for automorphisms of $II_1$ von Neumann algebras, \textit{Acta Math.} {\bf 134}(1975), 289-306.
\bibitem{CNT87} Connes A., Narnhoffer H., Thirring W., Dynamical entropy of $C^*$-algebras and von Neumann algebras, \textit{Commun. Math. Phys.} {\bf 112}(1987), 691-719.
\bibitem{fannes}
Fannes M., Nachtergaele B., Werner R. F., Ground states of VBS
models on Cayley trees, \emph{J. Stat. Phys.} {\bf 66} (1992)
939--973.
\bibitem{F.N.W} Fannes M., Nachtergaele B. Werner R. F., Finitely correlated
states on quantum spin chains, \textit{Commun. Math. Phys.} {\bf 144} (1992) 443--490.
\bibitem{Fid01} Fidaleo F., Fermi Markov states, \textit{J. Operator Theor.} {\bf 66}(2011), 385-414.
\bibitem{FM} Fidaleo F., Mukhamedov F., Diagonalizability of non homogeneous quantum Markov
states and associated von Neumann algebras, \textit{Probab. Math. Stat.} {\bf 24} (2004), 401-418.
\bibitem{Go98} Golodets, V., Neshveyev, S. Non-Bernoullian Quantum K-Systems, \textit{Comm Math Phys} {\bf 195}(1998), 213–232.
\bibitem{GZ} Golodets V.Y., Zholtkevich G.N. Markovian KMS states, \textit{Theor. Math.
Phys.} {\bf 56}(1983), 686-690.
\bibitem{Ki} Kieffer, J.C., A generalized Shannon–McMillan theorem for the action of an amenable group on a probability
space. \textit{Ann. Probab.} {\bf 3}(1975), 1031–1037.
\bibitem{Mo06} Moriya H., Markov property and strong additivity of von Neumann entropy for graded quantum
systems, \textit{J. Math. Phys.} {\bf 47}(2006), 033510.
%\bibitem{M04} Mukhamedov F., On factor associated with the unordered
%phase of $\lambda$-model on a Cayley tree. {\it Rep. Math. Phys.}
%{\bf 53}(2004), 1--18.
\bibitem{MBS161} Mukhamedov F., Barhoumi A., Souissi A., Phase transitions for quantum Markov chains associated with Ising type models on a Cayley tree, \textit{J. Stat. Phys.} {\bf 163} (2016), 544--567.
\bibitem{MBS162} Mukhamedov F., Barhoumi A., Souissi A., On an algebraic property of the disordered phase of the Ising model with competing interactions on a Cayley tree, \textit{Math. Phys. Anal. Geom.} {\bf 19}(2016), 21.
\bibitem{MR2004} Mukhamedov F., Rozikov U., On Gibbs measures of models with
competing ternary and binary interactions and corresponding von
Neumann algebras. \textit{J. Stat. Phys.} {\bf 114}(2004), No.
3-4, 825-848.
\bibitem{MS19}
Mukhamedov F., Souissi A., Quantum Markov States on Cayley trees,
\textit{J. Math. Anal. Appl.} {\bf 473}(2019), 313-333.
\bibitem{MS20} Mukhamedov, F., Souissi, A. Diagonalizability of Quantum Markov States on Trees. \textit{J Stat Phys} {\bf 182}(2021), Article 9.
\bibitem{MS21} Mukhamedov, F., Souissi A., Refinement of quantum Markov states on trees, \textit{J. Stat. Mech.} (2021) 083103
\bibitem{MS201} Mukhamedov F., Souissi A., Types of factors generated by
Quantum Markov Chains of Ising model with competing interactions
on the Cayley tree, \textit{Infin. Dimen. Anal. Quantum Probab. Related Topics}, {\bf 23}(2020), 2050019.
\bibitem{MSH21} Mukhamedov, F., Souissi, A. Hamdi, T. Quantum Markov Chains on Comb Graphs: Ising Model. \textit{Proc. Steklov Inst. Math.} {\bf 313}(2021), 178–192.
\bibitem{OW19} Ohmura K., Watanabe N., Quantum dynamical mutual entropy based on AOW entropy, \textit{Open Sys.
\& Infor. Dyn.} {\bf 26} (2019), 1950009.
\bibitem{Ohn04} Ohno H., Translation-Invariant Quantum Markov States, \textit{Interdiscip. Infor. Sci.} {\bf 10}(2004), 53–58.
\bibitem{OhyaPetz} Ohya, M. Petz, D., \textit{Quantum Entropy and Its Use}, Springer-Verlag, New York, 1993.
\bibitem{Or} Orus R.
A practical introduction of tensor networks: matrix product states and projected entangled pair states,
\textit{Ann of Physics} {\bf 349} (2014) 117--158.
\bibitem{Park94} Park, Y.M. Dynamical entropy of generalized quantum Markov chains. \textit{Lett Math Phys} {\bf 32}(1994), 63–74.
\bibitem{Park92} Park, Y. M. and Shin, H. H., Dynamical entropy of quasi-local algebras in quantum statistical mechanics, \textit{Comm. Math. Phys.} {\bf 144}(1992), 149–161.
\bibitem{Park} Park Y.M., Shin H.H., Dynamical entropy of generalized quantum Markov chains
over infinite dimensional algebras, \textit{J. Math. Phys.} {\bf 38}
(1997), 6287--6303.
\bibitem{Petz94} Petz, D. Entropy of Markov states. \textit{Math. Pura ed Appl.} {\bf 14}(1994), 33–42.
\bibitem{Petez88} Petz, D., Sufficiency of channels over von Neumann algebras, Q. J. Math. 39, 97–108, 1988
\bibitem{[RoOs96]}
Rommer S., Ostlund S.,
A class of ansatz wave functions for 1D spin systems and their relation to DMRG,
\textit{Phys. Rev. B} {\bf 55} (1997) 2164.
\bibitem{Roz2} Rozikov U.A. \textit{Gibbs measures on Cayley trees},
World Scientific, Singappore, 2013.
\bibitem{Sou21} Souissi A., A Class of Quantum Markov Fields on Tree-like Graphs: Ising-type Model on a Husimi Tree, \textit{Open Systems Information Dynamics} {\bf 28}(2021), 2150004.
\bibitem{Su97} Suyari H., Mean mutual entropy in quantum Markov chains, \textit{Electronics and Communications in Japan, Part 3}, {\bf 80}(1997), 104-112.
\bibitem{UmOhy84} Umegaki H., Ohya, M.: Quantum Entropies. Kyoritsu, Tokyo (1984).
%
%\bibitem{BP} Baker B.M., Powers R.T. Product states of certain
%group-invariant AF-algebras, \textit{J. Oper. Theor.} {\bf
%16}(1986), 3--50.
\bibitem{W2011} Watanabe, N. Note on entropies of quantum dynamical systems. Found. Phys. 41, 549–563 (2011)
\end{thebibliography}
\end{document}