On the automorphism group of the one-rooted binary tree (Q1372656)

Let \(T_2\) be the one-rooted regular binary tree and \(A\) its group of automorphisms. Denote by \(M=\{0,1\}^*\) the monoid of words on the set \(\{0,1\}\), ordered \((\geq)\) by the rule: \(u\geq v\) if \(u\) is a prefix for \(v\); \(T_2\) can be viewed as the diagram of this order \((M;\geq)\). Defining \(d(u,v)=| u|+| v|-2| w|\) for any two words \(u,v\in M\) with their largest common prefix equal to \(w\), gives a metric space \((M,d)\) and \(A\) can be considered as the group of all isometries of \(T_2\). Also, for any group \(G\) having an enumerable normal series \((G_i\mid i\geq 0)\) with \([G_i:G_{i+1}]=2\) and \(\bigcap_iG_i=(1)\) there exists a `coset tree' that can serve for rooted binary tree on which \(G\) acts faithfully. For an automorphism of \(T_2\) use the representation \(\cdots((x^{i_{00}},x^{i_{01}}),(x^{i_{10}},x^{i_{11}})) (x^{i_0},x^{i_1}) x^{i_\emptyset}\), with \(x\) denoting the automorphism of \(T_2\) given by the rule: for all \(m\in M\), \(0m\mapsto 1m\) and \(1m\mapsto 0m\); here \(i_m\in\{0,1\}\). An automorphism \(\alpha\) fixing the vertices labelled by 0 and 1, is written as \((\alpha_0,\alpha_1)\), \(\alpha_i\) denoting the automorphism of the subtree rooted in \(i\in\{0,1\}\). Denote by \(B\) the base group in \(A\), generated by the set \(\{x_0,x_1,x_2,\dots\}\), where \(x_0=x\), \(x_1=(1,x_0)\), \(x_2=((1,1),(1,x_0))\); i.e., \(x_i=(1,x_{i-1})\), \(i\geq 1\). It appears that \(B\) is the (restricted) infinitely iterated wreath product \((\dots\wr C_2)\wr C_2\). Starting from the observation that the elements of \(B\) determine finite-state automata, the authors define functionally recursive automorphisms and find some properties of the set of such automorphisms. In more details, let \(\Phi=\{\alpha,\dots,\delta\}\) be a finite subset in \(A\), \(\alpha=(\alpha_0,\alpha_1)\cdot x^{i_\emptyset(\alpha)}\),\dots, \(\delta=(\delta_0,\delta_1)\cdot x^{i_\emptyset(\delta)}\). Take the free group \(F\) generated by distinct symbols \(a,\dots,d\) -- one for every element in \(\Phi\) -- and consider the free product \(F*B\). The authors call the set \(\Phi\) functionally recursive if in \(F*B\) there are words \(A_0,A_1;\dots;D_0,D_1\) such that \(\alpha_i\) is obtained by substituting for \(a,\dots,d\) in \(A_i\) (\(i=0,1\)) the automorphisms \(\alpha,\dots,\delta\), correspondingly; \dots; \(\delta_i\) is obtained similarily from \(D_i\) (\(i=0,1\)). Now, an automorphism is called functionally recursive if it is an element of some functionally recursive set of automorphisms. The authors prove (Theorem 3.1) that the set of all such automorphisms of \(T_2\) is a subgroup in \(A\) and this countable set is strictly larger than the set of its finite-state automorphisms. So, there appears the link of this paper with questions of representing elements of certain Burnside groups; see \textit{S. V. Aleshin} [Mat. Zametki 11, 319-328 (1972; Zbl 0246.20024)], \textit{R. I. Grigorchuk} [Funkts. Anal. Prilozh. 14, No. 1, 53-54 (1980; Zbl 0595.20029)] and \textit{N. Gupta, S. Sidki} [Contemp. Math. 33, 232-246 (1984; Zbl 0553.20016)]. Further, the authors have investigated the base group \(B\) in terms of invariant subgroups, conjugacy classes and endomorphisms induced by conjugation by elements of \(A\). Among the many results of the authors there are the following: (Theorem 4.1) that a Krull-Schmidt theorem holds for \(B\); (Theorem 7.1) that \(\Aut(B)\cong N_A(B)\); and (Proposition 7.10) that \(\Aut(B)\) contains a copy of \(\Aut(T_2)\). For any element \(b\in B\) consider the element \(\alpha\in A\) recursively defined by \(\alpha=(\alpha\cdot b,\alpha)=\alpha^{(1)}(b,1)\); here \(\alpha=\dots(b,1)^{(2)}\cdot(b,1)^{(1)}\cdot(b,1)\). It appears (Theorem 8.1) that such elements \(\alpha\in A\) generate a subsemigroup \(S\) of \(\text{End}_A(B)\) under multiplication, such that \(S^{-1}\cap\text{End}_A(B)=(1)\). It is shown as well that \(N_A(B)\) is properly contained in \(\text{End}_A(B)\). The study of \(\Aut(T_2)\) in this paper is based on the techniques used by \textit{S. Sidki} [J. Algebra 110, 24-55 (1987; Zbl 0623.20025)].