On infinite Steiner systems (Q1183984)

A Steiner system \(S(t,k,v)\) consists of a set \(V\) (called the base set) of cardinality \(v\) and a family \({\mathcal B}\) of \(k\)-element subsets of \(V\) (called blocks) such that every \(t\)-element subset of \(V\) occurs in precisely one block. Here \(t\) and \(k\) are positive integers with \(t\leq k\). Almost all of the vast literature which now exists on Steiner systems (for a bibliography see \textit{J. Doyen} and \textit{A. Rosa} [Ann. Discrete Math. 7, 317-349 (1980; Zbl 0442.51006)]) deals with the case where \(V\) is finite. In this case no Steiner systems have yet been constructed with \(t>5\) and only a finite number are known with \(t>3\). However, in the case where \(v=\aleph_ 0\), \textit{E. Köhler} [J. Geometry 9, 73-77 (1977; Zbl 0345.50011)] has proved the existence of an \(S(t,k,\aleph_ 0)\) for all \(t\), \(k\) satisfying \(1<t<k\). In the book by \textit{T. Beth}, \textit{D. Jungnickel} and \textit{H. Lenz} [Design theory (1985; Zbl 0569.05002)], it is observed that the proof is nonconstructive and they remark that ``the explicit description of infinite \(t\)-designs appears difficult, perhaps hopeless'' (p. 374). In this paper we give a direct construction for systems \(S(t,t+1,v)\) in the cases where \(t\) is a positive integer and \(v=\aleph_ 0\), or \(2^{\aleph_ 0}\), or \(2^ \alpha\) where \(\alpha\) is any infinite cardinal. The construction yields what is termed a large set of mutually disjoint \(S(t,t+1,v)\) systems in each of these cases --- i.e. in each case we obtain a family \(\{S_ v\}\) of infinite Steiner systems such that every \((t+1)\)-element subset of the base set appears in precisely one Steiner system of the family. The method extends the results of the authors in [Ars. Comb. 24B, 189-216 (1987; Zbl 0724.05009)] regarding countably infinite Steiner triple systems \(S(2,3,\aleph_ 0)\).