Abstract:The Minkowski problem in Gaussian probability space is studied in this paper. In addition to providing an existence result on a Gaussian-volume-normalized version of this problem, the main goal of the current work is to provide uniqueness and existence results on the Gaussian Minkowski problem (with no normalization required).
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is the Minkowski problem in Gaussian probability space. Specifically, the paper explores, given a finite Borel measure \(\mu\), what necessary and sufficient conditions exist such that there is a convex body \(K\) containing the origin in its interior and whose Gaussian surface area measure \(S_{\gamma_n, K}\) is exactly equal to the given measure \(\mu\). Moreover, if such a \(K\) exists, to what extent is it unique.
The key issues mentioned in the paper can be summarized as follows:
1. **Existence**: For a given finite Borel measure \(\mu\), does there exist a convex body \(K\) such that \(S_{\gamma_n, K}=\mu\)?
2. **Uniqueness**: If such a convex body \(K\) exists, is it unique? In particular, when \(\gamma_n(K)\geq1/2\), can uniqueness be guaranteed?
3. **Normalization problem**: In some cases, can a solution be found by the normalization method, that is, does there exist a constant \(c\) and a convex body \(K\) such that \(\mu = cS_{\gamma_n, K}\)?
### Main Results
1. **Uniqueness Results**:
- **Theorem 1.1**: Suppose \(K\) and \(L\) are two convex bodies containing the origin in their interiors and they both solve the Gaussian Minkowski problem, that is, \(S_{\gamma_n, K}=S_{\gamma_n, L}=\mu\). If \(\gamma_n(K),\gamma_n(L)\geq1/2\), then \(K = L\).
2. **Solution to the Normalization Problem**:
- **Theorem 1.2**: Suppose \(\mu\) is a finite even Borel measure that is not concentrated in any closed hemisphere. Then for each \(0 < \alpha<\frac{1}{n}\), there exists a convex body \(K\) symmetric about the origin such that \(\mu = cS_{\gamma_n, K}\), where \(c=\frac{1}{\gamma_n(K)^{1 - \alpha}}\).
3. **Existence of Non - Normalized Problem**:
- **Theorem 1.3** (Existence of Smooth Solutions): Let \(0 < \alpha< 1\) and \(f\in C^{2,\alpha}(S^{n - 1})\) be a positive even function satisfying \(\|f\|_{L^1}<\frac{1}{\sqrt{2\pi}}\). Then there exists a unique \(C^{4,\alpha}\) convex body \(K\) symmetric about the origin such that \(\gamma_n(K)>1/2\) and satisfies the equation \(\frac{1}{(\sqrt{2\pi})^n}e^{-|\nabla h_K|^2+\frac{h_K^2}{2}}\det(\nabla^2 h_K + h_KI)=f\).
- **Theorem 1.4** (Existence of Weak Solutions): Let \(\mu\) be a measure symmetric about the origin, not concentrated in any subspace, and \(\|\mu\|<\frac{1}{\sqrt{2\pi}}\). Then there exists a unique convex body \(K\) symmetric about the origin such that \(\gamma_n(K)>1/2\) and \(S_{\gamma_n, K}=\mu\).
These results not only solve the existence and uniqueness of the Minkowski problem in Gaussian probability space but also provide an important theoretical basis for further research in this field.