Abstract:We introduce a random barrier to a supercritical branching random walk in an i.i.d. random environment $\{\mathcal{L}_n\}$ indexed by time $n,$ i.e., in each generation, only the individuals born below the barrier can survive and reproduce. At generation $n$ ($n\in\mathbb{N}$), the barrier is set as $\chi_n+\varepsilon n,$ where $\{\chi_n\}$ is a random walk determined by the random environment. Lv \& Hong (2024) showed that for almost every $\mathcal{L}:=\{\mathcal{L}_n\},$ the quenched survival probability (denoted by $\varrho_{\mathcal{L}}(\varepsilon)$) of the particles system will be 0 (resp., positive) when $\varepsilon\leq 0$ (resp., $\varepsilon>0$). In the present paper, we prove that $\sqrt{\varepsilon}\log\varrho_\mathcal{L}(\varepsilon)$ will converge in Probability/ almost surely/ in $L^p$ to an explicit negative constant (depending on the environment) as $\varepsilon\downarrow 0$ under some integrability conditions respectively. This result extends the scope of the result of Gantert et al. (2011) to the random environment case.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the asymptotic behavior of the quenched survival probability of supercritical branching random walks with obstacles (BRWre) in a random environment. Specifically, the author studies the limit behavior of the product of the logarithm of the quenched survival probability \(\rho_L(\varepsilon)\) and \(\sqrt{\varepsilon}\) as the obstacle parameter \(\varepsilon\) tends to 0. The goal of the paper is to prove that under certain integrability conditions, \(\sqrt{\varepsilon}\log\rho_L(\varepsilon)\) converges in probability, almost surely, or in the \(L^p\) sense to an explicit negative constant.
### Problem Background
- **Model**: The paper considers branching random walks (BRWre) in a time - inhomogeneous independent and identically distributed random environment. In each generation, only individuals whose birth positions are below the obstacle can survive and reproduce.
- **Obstacle**: In the \(n\) - th generation, the obstacle is set as \(\chi_n+\varepsilon_n\), where \(\{\chi_n\}\) is a random walk determined by the random environment.
- **Known Results**: Lv and Hong (2024) proved that for almost every random environment \(L\), when \(\varepsilon\leq0\), the quenched survival probability \(\rho_L(\varepsilon) = 0\); when \(\varepsilon>0\), the quenched survival probability is positive.
### Main Contributions of the Paper
- **Main Results**: The paper proves that under certain integrability conditions, \(\sqrt{\varepsilon}\log\rho_L(\varepsilon)\) will converge in probability, almost surely, or in the \(L^p\) sense to an explicit negative constant depending on the environment.
- **Methods**: The author uses Mogul'skiĭ estimates in a random environment (a small - deviation principle for random walks), combined with probability theory tools such as measure transformation, first - and second - moment methods, and the small - deviation principle.
- **Extensions**: This result extends the results of Gantert et al. (2011) in a deterministic environment to make them applicable to a random environment.
### Specific Conclusions
1. **Convergence in Probability**: If assumption (A1) holds, then for any \(x > 0\), \(\lim_{\varepsilon\downarrow0}P(\sqrt{\varepsilon}\log\rho_L(\varepsilon)>\gamma + x)=0\).
2. **Almost Sure Convergence**: If assumption (A1 +) holds, then \(\lim_{\varepsilon\downarrow0}\sqrt{\varepsilon}\log\rho_L(\varepsilon)\leq\gamma\) almost surely.
3. **\(L^p\) Convergence**: If assumptions (A1), (A2) and (A3) hold, then for any \(p\in[0,\min(\lambda_8,\nu_0)- 2)\), \(\sqrt{\varepsilon}\log\rho_L(\varepsilon)\to\gamma\) in the \(L^p\) sense.
### Technical Difficulties
- **Auxiliary Process**: To prove the above results, the author constructs an auxiliary supercritical branching process and studies the lower - bound estimate of its survival probability.
- **Complexity**: Due to the existence of the random environment, estimating the quenched survival probability becomes more complex, and it is necessary to further study the survival probability of BRWre with obstacles within a specific time.
### Future Prospects
- **N - BRWre**: Future research can consider whether certain types of N - BRWre exhibit Brunet - Derrida behavior, that is, the maximum and minimum displacements.