On structure of topological entropy for tree-shift of finite type (Q2034033)

The authors investigate the topological entropy for hom Markov shifts \(\mathcal{T}_M\) on \(d\)-trees, thus extending a theorem by \textit{D. Lind} and \textit{B. Marcus} [An introduction to symbolic dynamics and coding. Cambridge: Cambridge University Press (1995; Zbl 1106.37301)] to the case where \(S\) is the family of hom Markov tree-shifts. Let \(A, B\in \{0, 1\}^{k\times k}.\) It is proved that if \(B\) is irreducible and \(A >B\), then \(h(\mathcal{T}_A)>h(\mathcal{T}_B)\). For any \(a, b\geq 1\) and \(0\leq l \leq b\), consider the reducible matrix \[M=M(a,b;l)=\left[ \begin{array}{cc} E_a & R_l \\ O & E_b \end{array}\right],\] where \(O\) is \(b \times a\) zero matrix and \(R_l\in \{0, 1\}^{a\times b}\) has the same row sum \(l\). \(E_a\) and \(E_b\) are the irreducible components of the matrix \(M\). For \(b > a\), the authors provide a classification result related to the equality \(h(\mathcal{T}_{M(a,b;l)}) = \log b\). Then they extend the results valid for reducible matrices \(M\) with two irreducible components to some more general cases. If \(M\) is a reducible adjacency matrix with \(q\) irreducible components \(M_1,\dots, M_q\), the authors prove that \(h(\mathcal{T}_M) = \max_{1\leq i\leq q} h(\mathcal{T}_{M_i})\) fails generally. In Section 4, they deal with the relative denseness of the sets \[\mathcal{H}^{(d)} = \{h(\mathcal{T}_{M}) : \mathcal{T}_{M}\ \text{is on \(d\)-tree and \(M\) is binary}\},\] and \[ \mathcal{H}^{(d)}_{irr} = \{h(\mathcal{T}_{M}) : \mathcal{T}_{M}\ \text{is on \(d\)-tree, and \(M\) is binary and irreducible}\},\] in the set of entropies of all hom tree-shifts. Then they study the denseness of \(\mathcal{H}^{(d)}_{irr} \), or equivalently the denseness of \(\mathcal{H}^{(d)}\). In the last section, further discussions and open problems are presented.