Quenched central limit theorems for random walks in random scenery (Q1939343)

Let \(d\geq 1\) be an integer. Let \(\omega= (\omega_x)_{x\in\mathbb{Z}^d}\) be a sequence of i.i.d. random variables defined on a probability space \((\Omega,{\mathcal F},\operatorname{P})\). Let \((X_n)_{n\geq 1}\) be a sequence of i.i.d. random variables defined on another probability space \((\Omega',{\mathcal A},\operatorname{P}')\), taking values in \(\mathbb{Z}^d\). Define the random walk \(S= (S_n)_{n\geq 0}\) by \(S_0= 0\) and \(S_n= \sum^n_{k=1} X_k\) \((n\geq 1)\). We assume that, each time, the random walk evolves in \(\mathbb{Z}^d\), i.e., the walk is truely \(d\)-dimensional. Define the random walk in random scenery by \(Z_n= \sum^n_{k=1} \omega_{S_k}\). We assume that \(\omega_0=^{(\mathrm{law})}-\omega_0\), \(\operatorname{E}|\omega_0|^k< \infty\) \((k\geq 1)\), and \(\operatorname{E}\omega^2_0= 1\). In this paper, the authors establish quenched central limit theorems in the following three cases. (i) \((d=1)\) Suppose that the support of \(X_1\) is an aperiodic set of \(\mathbb{Z}^+\), \(m= \operatorname{E}X_1<\infty\) and \(\operatorname{E}X^2_1<\infty\). Then \[ ([Z_n- \operatorname{E}Z_n]/\sqrt{n})_{n\geq 1}@>(\mathrm{law})>>{\mathcal N}(0,1- m^{-1}),\;\omega\text{-a.s.} \] (ii) \((d=2)\) Suppose that the random walk is symmetric with finite nonsingular covariance matrix \(\Sigma\). Then, for all \(\nu> 0\), \[ Z_{t_m}/\sqrt{t_m\log t_m}@>(\mathrm{law})>>{\mathcal N}(0, \sigma^2), \] where \(t_m= [\exp(m^{1+\nu})]\) and \(\sigma^2= (2\pi\sqrt{\det\Sigma})^{-1}\). (iii) Suppose the random scenery is centered, square integrable with variance one and that one of the following assumptions holds: (1) \(d\geq 3\) and \(X_1\) is square integrable; (2) \(X_1\) is in the domain of attraction of a strictly stable law with index \(\alpha\in(0, 2)\) and \(d>\alpha\). Then \[ (Z_{[nt]}/\sqrt{n})_{t\geq 0}@>(\mathrm{law})>>(\sigma B_t)_{t\geq 0}, \] where \((B_t)_{t\geq 0}\) is a real Brownian motion and \[ \sigma^2= \gamma^2 \sum_{k\geq 0} k^2(1-\gamma)^{k- 1} \] with \(\gamma= \operatorname{P}' (S_k\geq 0\text{ for any }k\geq 1)\in (0,1)\).