Infinitely divisible measures on p-adic groups (Q2277647): Difference between revisions

The paper is a contribution to the ``embedding problem'' for totally disconnected locally compact groups. The problem can be stated as follows: Let \(\mu\) be an infinitely divisible probability measure on a group G, is \(\mu\)- or a suitable shift \(\epsilon_ x*\mu\)- embeddable into a continuous convolution semigroup? While for (real) Lie groups nowadays the embedding problem is nearly solved [see the recent publications of \textit{S. G. Dani} and \textit{M. McCrudden}, e.g. Math. Z. 199, No.3, 369-385 (1988)], for (non-Abelian) groups our knowledge is less satisfactory. In the paper under review the author solves the problem for p-adic algebraic groups: If \(\mu\) is infinitely divisible on a closed subgroup of \(Gl_ n({\mathbb{Q}}_ p)\), then there exist a unipotent element x in the centralizer of the support of \(\mu\), such that the shift \(\epsilon_ x*\mu\) is continuously embeddable. The proof follows the main ideas developed during the last years, the tools - e.g. root compactness, submonogenous embedding, continuous embedding - are slightly generalized and suitable adapted to be used in the case of p- adic groups. Since on p-adic groups point measures are not continuously embeddable, the main result is in some sense the best possible.