The Gauss image problem for pseudo-cones

From MaRDI portal





Let \(C\) be a closed and convex cone \(C\) in the \(n\)-dimensional Euclidean space \(\mathbb{R}^n\), \(n\ge2\); assume that \(C\) has non-empty interior and does not contain any line. A \(C\)-pseudo-cone is a nonempty closed convex subset \(K\) of \(\mathbb{R}^n\), not containing the origin, such that \(\lambda K\subset K\) for every \(\lambda\ge1\), and such that the recession cone of \(K\), given by \(\{z\in\mathbb{R}^n\colon K+z\subset K\}\), coincides with \(C\). Let \N\[\N\Omega_C=C\cap\mathbb{S}^{n-1},\quad \Omega_{C^\circ}=C^\circ\cap\mathbb{S}^{n-1},\N\]\Nwhere \(\mathbb{S}^{n-1}\) is the unit sphere of \(\mathbb{R}^n\), and \(C^\circ\) is the dual cone of \(C\).\N\NThe radial Gauss image of a \(C\)-pseudo-cone \(K\) at a set \(\eta\subset\Omega_C\), denoted by \(\boldsymbol{\alpha}_K(\eta)\), is defined as the set of all outer unit normal vectors to \(K\) at points of the form \(rv\), where \(v\in\eta\) and \(r>0\). The set \({\boldsymbol\alpha}_K(\eta)\) is contained in the closure of \(\Omega_{C^\circ}\), for every \(\eta\subset\Omega_C\).\N\NConsider a finite, non-zero measure \(\lambda\), defined on Borel measurable subsets of the closure of \(\Omega_{C^\circ}\), vanishing on subsets of dimension \((n-2)\). For Borel sets \(\eta\subset\Omega_C\), let \N\[\N\lambda(K,\eta)=\lambda(\boldsymbol{\alpha}_K(\eta)).\N\]\NThe author considers the following Gauss image problem for pseudo-cones: given a Borel measure \(\mu\) on \(\Omega_C\), find necessary and sufficient conditions on \(\lambda\) and \(\mu\) such that there exists a \(C\)-pseudo-cone \(K\) verifying: \N\[\N\lambda(K,\cdot)=\mu(\cdot). \N\]\NThe main result of the article is the following theorem.\N\NTheorem. Let \(\lambda\) be a measure defined on the Lebesgue measurable subsets of the closure of \(\Omega_C\) and assume that it is non-zero, finite, and zero on Borel sets of Hausdorff dimension \((n-2)\). Let \(\mu\) be a Borel measure on \(\Omega_C\). There exists a \(C\)-pseudo-cone \(K\) with \N\[\N\mu(\cdot)=\lambda(K,\cdot) \N\]\Nif and only if \(\lambda(\Omega_{C^\circ})=\mu(\Omega_C)\).











This page was built for publication: The Gauss image problem for pseudo-cones

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6899305)