scientific article; zbMATH DE number 1498458
This article is a continuation of the authors' article [ibid. 2, No. 2, 57-97 (1999; Zbl 0946.60040)]. Let \(\gamma,\xi(2),\xi(3),\dots\) be independent random vectors in \({\mathbb R}^2\) with nonzero absolutely continuous component and satisfying the Cramér condition; the vectors \(\xi(i)\), \(i=2,3,\dots\), are assumed to be identically distributed. Let \(S(\gamma,n)\) be a random walk with initial distribution \(\gamma\), i.e., \(S(\gamma,n)=\gamma+\xi(2)+\cdots+\xi(n)\). Put \(\eta_-=\inf\{n\geq 1:S_2(\gamma,n)\leq 0\}\) and, for \(y\geq 0\), \(\eta_-(-y)=\inf\{n\geq 1:S_1(\gamma,n)\leq -y\}\). For every fixed set \(A\subset(-\infty,0]\), asymptotic representations for the probability \({\mathbb P}\{S_1(\eta_-)>t\), \(S_2(\eta_-)\in A\), \(\eta_-(-y)>\eta_-\}\) are deduced, as \(t\to\infty\). Let \[ \mu=\sup\{\lambda_1: {\mathbb E}e^{\lambda_1\xi_1(2)+\lambda_2\xi_2(2)}\leq 1 \text{ for some }\lambda_2\in{\mathbb R}\}. \] The main result of the article asserts that, for all sufficiently small \(\lambda>0\), \[ {\mathbb E}\{e^{\lambda S_2(\eta_-)};\;S_1(\eta_-)>t, \;\eta_-(-y)>\eta_-\} \sim ce^{-t\mu}t^{-3/2}\text{ as }t\to\infty, \] where the constant \(c\) depends on \(\lambda\) and \(y\).
- Exit probability of two-dimensional random walk from the quadrant
- Random walks in the positive quadrant. II. Integral theorem
- Random walks in the positive quadrant. III: Constants in integral and local theorems
- Taming reluctant random walks in the positive quadrant
- Positivity of integrated random walks
- Exit probability of two-dimensional random walk from the quadrant
- scientific article; zbMATH DE number 1438903 (Why is no real title available?)
- Escape time for a random walk from an orthant
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4499530)