Cones in numerical modules (Q757777)

Cones in numerical modules, i.e. modules over a subring of the real numbers, generalize convex cones (in vector spaces) as well as semigroups (in abelian groups). A cone P in a numerical module with archimedean element a is called a Stone cone (following D. W. Dubois) if \(a+nx\in P\) for all \(n\in {\mathbb{N}}\) implies \(x\in P\). It is shown that a cone is a Stone cone iff it coincides with its infinitesimal hull or, equivalently, iff it is determined by its dual cone. When finitely generated, a cone is a Stone cone iff it is integrally closed and generates the module properly. Also available is a characterization in terms of a Hilbert basis of the module. Applying the above characterizations, it is shown that for Stone cones a generalized Farkas lemma does hold as well as an extension of a theorem of C. Davis by which a cone can be represented as the intersection of a standard cone and a submodule.