Abstract:In this work, we study the random metric for the critical long-range percolation on $\mathbb{Z}^d$. A recent work by Bäumler [3] implies the subsequential scaling limit, and our main contribution is to prove that the subsequential limit is uniquely characterized by a natural list of axioms. Our proof method is hugely inspired by recent works of Gwynne and Miller [42], and Ding and Gwynne [25] on the uniqueness of Liouville quantum gravity metrics.
What problem does this paper attempt to address?
The key problem that this paper attempts to solve is to prove that in the Critical Long - Range Percolation (LRP) model, the scale limit of the chemical distance exists and is unique. Specifically, the main contribution of the paper lies in uniquely characterizing this scale limit through a set of natural axioms, which can describe the properties of the random metric in the critical long - range percolation model.
### Background and Problem of the Paper
The critical long - range percolation model is a percolation model on \( \mathbb{Z}^d \). In this model, each pair of vertices can be connected by an edge. Specifically, for the nearest - neighbor edge \( \langle i, j \rangle \) (that is, \( \|i - j\|_1 = 1 \)), the probability of its occurrence is \( p_1 \); while for the long - distance edge \( \langle i, j \rangle \) (that is, \( \|i - j\|_1>1 \)), its probability of occurrence is:
\[ 1 - \exp\left( -\beta \int_{V_1(i)} \int_{V_1(j)} \frac{1}{|u - v|^{2d}} \, du \, dv \right) \]
where \( V_1(i) \) is a cube centered at \( i \) with a side length of 1, \( \| \cdot \|_1 \) is the \( \ell_1 \)-norm, and \( | \cdot | \) is the \( \ell_2 \)-norm.
### Main Results
The main result of the paper is to prove that the scale limit of the chemical distance \( b_d \) exists and is unique. Specifically, let \( b_d \) be the chemical distance in the discrete \( \beta \)-LRP model, and \( b_a(n) \) be the median of \( b_d(0, n1) \), where \( 1=(1, 1, \ldots, 1)\in \mathbb{R}^d \). Define \( b_D_n(x, y)=b_a(n)^{-1}b_d(\lfloor nx \rfloor, \lfloor ny \rfloor) \). Then there exists a unique random metric \( D \) such that \( b_D_n \) converges in distribution to \( D \), and \( D \) satisfies the continuity under the locally uniformly convergent topology.
### Axiomatic Characterization
To prove this result, the author introduced a set of axioms to uniquely characterize the strong \( \beta \)-LRP metric \( D \):
1. **Length Space**: Consider the equivalence relation \( \sim \) defined as \( x \sim y \) if and only if \( D(x, y) = 0 \). Then \( (\mathbb{R}^d / \sim, D) \) forms a length space, and there exists a geodesic between any two elements \( x, y\in \mathbb{R}^d \).
2. **Locality**: Let \( U \) be a deterministic open subset of \( \mathbb{R}^d \). Then the internal metric \( D(\cdot, \cdot; U) \) of \( D \) inside \( U \) is determined by the restriction of \( E \) on \( U\times U \) and is independent of all edges with at least one endpoint outside \( U \).
3. **Regularity**: For any \( x, y\in \mathbb{R}^d \), if \( \langle x, y \rangle\in E \), then \( D(x, y) = 0 \).
4. **Translation and Scale Covariance**: For any \( r > 0 \) and \( z\in \mathbb{R}^d \), we have \( D(r\cdot+z, r\cdot