Abstract:In this work we are concerned with the crossover between strong disorder (SD) and weak disorder (WD) behaviors in three well-known problems that involve minimal paths: directed polymers (directed paths with fixed starting point and length), optimal paths (undirected paths with fixed end-to-end/spanning distance) and undirected polymers (undirected paths with fixed starting point and length). We present a unified theoretical framework from which we can easily deduce the scaling of the crossover point of each problem. Our theory is based on the fact that the SD limit behavior of these systems is closely related to the corresponding percolation problem. As a result, the properties of those minimal paths are completely controlled by the so-called red bonds of percolation theory. Our model is first addressed numerically and then approximated by a ``two-term'' approach. This approach provides us with an analytical expression that seems to be reasonably accurate. The results are in perfect agreement with our simulations and with most of the results reported in related works. Our research also lead us to propose this crossover point as a universal measure of the disorder strength in each case. Interestingly, that measure depends on both the statistical properties of the disorder and the topological properties of the network.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is a unified theoretical framework for the crossover between optimal paths and directed/undirected polymers from strong - disorder (SD) behavior to weak - disorder (WD) behavior in disordered media. Specifically, the paper focuses on the behavioral changes of these paths during the transition from strong - disorder to weak - disorder states in three classical problems involving minimal paths, namely, directed polymers (directed paths with a fixed starting point and length), optimal paths (undirected paths with a fixed end - to - end distance), and undirected polymers (undirected paths with a fixed starting point and length).
### Main contributions of the paper
1. **Unified theoretical framework**:
- Proposed a unified theoretical framework from which the scaling behavior of the crossover point for each problem can be easily derived.
- The theory is based on the close relationship between the strong - disorder limit behavior and the corresponding percolation problem, that is, the properties of the minimal paths in these systems are completely controlled by the "red bonds" in percolation theory.
2. **Numerical simulation and analysis**:
- Verified the theoretical model through numerical simulation and proposed a "two - term" method to provide analytical expressions. The results are highly consistent with those in simulations and other related works.
3. **Universality measure**:
- Proposed a method for measuring the disorder strength of universality, which depends on the statistical characteristics of the disorder and the topological characteristics of the network.
### Specific problem description
- **Background**: The study of minimal paths in disordered systems has been going on for decades and has a wide range of applications, including polymer science, transport, fluid flow in porous media, human behavior, social networks, communication networks, and traffic engineering.
- **Disordered system model**: A disordered system is usually modeled as a regular lattice or a random network, and its edges are assigned a positive random weight \( w \), which is usually considered as an independent and identically distributed (i.i.d.) random variable with a common probability density \( f(w) \) and a cumulative distribution \( F(w) \).
- **Minimal path**: The minimal path refers to the path with the smallest sum of weights on the path, and the total weight is denoted as \( W_{\text{opt}} \). The optimization process is carried out on the set of paths that meet the geometric constraints, which determine the type of problem.
### Three main problems
1. **Directed polymer (DPRM)**:
- The starting point and the path length are fixed, but the end point is not fixed.
- In the \((D + 1)\)-dimensional DPRM, the lattice edges are directed in the selected direction (usually called time \( t \)), where \( t \) represents the length of the polymer, and the dimension \( D \) of the transverse hyperplane is usually called space \( x \).
2. **Optimal path (OP)**:
- It includes the optimal path on a regular lattice (OPRL) and the optimal path on a random network (OPRN).
- In the \( D \)-dimensional OPRL, consider the optimal path connecting the opposite sides of a \( D \)-dimensional lattice, or consider the optimal path with a distance \( r \) between two fixed points.
- In the OPRN problem, consider a random network with a size of \( N \) (the number of nodes), and \(\ell_{\text{opt}}\) represents the average length of the optimal path between two nodes in the network.
3. **Undirected polymer (UPRM)**:
- It is the undirected analogue of DPRM.
- In the \( D \)-dimensional UPRM, consider a polymer with a fixed length \(\ell\) fixed at one end in the center of a \( D \)-dimensional lattice, whose configuration is self - avoiding walk (SAW), with a fixed starting point and length, and a free end point.
### Disorder effects
- **Weak disorder (WD)**: Almost all links contribute to \( W_{\text{opt}} \).
- **Strong disorder (SD)**: The total weight of the path is mainly determined by the maximum weight \( w_{\text{max}}\) on the path, that is, \( W_{\text{opt}}=w_{\text{max}}\) (called the maximum principle).
- **Crossover point**: The paper focuses on the crossover point from strong disorder to weak disorder, that is, below this point, the minimal path behaves...