Cone-volume measure of general centered convex bodies (Q890138)

Given a convex polytope \(K\subset \mathbb{R}^n\) with facets \(F_1,\dots,F_m\) and corresponding outer unit normals \(u_1,\dots,u_m\), its surface area measure \(S_K(\cdot)\) is the Borel measure on the Euclidean unit sphere \(S ^{n-1}\subset \mathbb{R}^n\) which is given by \(S_K(\cdot)=\sum_{i=1}^m V_{n-1}(F_i)\delta_{u_i}(\cdot)\). Here \(\delta_{u_i}\) is the Dirac delta measure on \(S^{n-1}\) concentrated at \(u_i\), and \(V_{n-1}(F_i)\) is the \((n-1)\)-dimensional volume of \(F_i\). The surface area measure of a convex body \(K\) is the weak limit of the surface area measures of some sequence of polytopes converging to \(K\). The Minkowski problem asks which measures on \(S^{n-1}\) are surface area measures of convex bodies. The cone-volume measure of a polytope \(K\) is defined similarly to the surface area measure, but using the volumes of the cones with apex the origin \(o\) and basis \(F_i\), instead of \(V_{n-1}(F_i)\). This concept can also be extended to convex bodies, and the logarithmic Minkowski problem asks which measures on \(S^{n-1}\) are cone-volume measures of convex bodies with \(o\) in its interior. The authors prove that the cone volume measure \(\mu\) of a convex body \(K\) with centroid at the origin satisfies the subspace concentration condition for any linear subspace \(L\subseteq \mathbb{R}^n\), namely \(\mu(L\cap S^{n-1})\leq \frac{\dim L}{n} \mu(S^{n-1}).\) This extends the similar former result for centered polytopes and for the origin symmetric convex bodies. It is also proved that for every centered convex body, the cone-volume of any open hemisphere is greater than or equal to \(\frac{1}{2n}\) times the volume of the convex body. The equality holds if and only if \(K\) is a cylinder whose generating segment is parallel to the center of the hemisphere.