Extremum problems for the cone volume functional of convex polytopes (Q607344): Difference between revisions

In 2001, Lutwak, Yang and Zhang introduced the cone volume functional \(U(P)\) of a convex polytope in \(n\)-dimensional euclidean space as \[ U(P)^n=\frac 1{n^n} \sum_{u_{i_1}\wedge \dots \wedge u_{i_n}\neq 0} h_{u_{i_1}} \cdots h_{u_{i_n}}a_{u_{i_1}}\cdots a_{u_{i_n}}, \] where \(u_1, \dots, u_N\) are the outer normal unit vectors to the corresponding facets \(F_1, \dots,F_N\) of \(P\), and the facet with outer normal vector \(u_i\) has area \(a_i\) and distance \(h_i\) from the origin. The cone volume functional is an effective tool to study Schneider's projection problem. One fundamental -- but still open -- extremum problem, on the ratio of \(U\) to \(V\) was posed by Lutwak, Yang and Zhang: Conjecture. If \(P\) is a convex polytope in \(n\)-dimensional euclidean space with its centroid at the origin, then \[ \frac {U(P)}{V(P)}\geq \frac{(n!)^{\frac 1n}}n, \] with equality if and only if \(P\) is a parallelotope. The first progress on this conjecture was attributed to He, Leng and Li. In 2006 they proved that the conjecture is true when restricted to the class of origin-symmetric convex polytopes. The present paper is devoted to the study of the above conjecture. The author gives affirmative answer to the conjecture for dimensions 2 and 3. He also uses techniques of geometric symmetrization and methods of projection to tackle the conjecture, which are readily applicable to symmetric convex polytopes. As a consequence he provides a mostly simplified proof of the conjecture for the case of origin-symmetric convex bodies. He also obtains some new sharp inequalities characterizing parallelotopes which are closely related to the functional \(U\). He remarks that it is surprising to see that parallelotopes are the only minimizers of the functional \(U\).