The asymptotics of stable sausages in the plane (Q1184078)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | The asymptotics of stable sausages in the plane |
scientific article |
Statements
The asymptotics of stable sausages in the plane (English)
0 references
28 June 1992
0 references
\textit{J. -F. Le Gall} [J. Funct. Anal. 88, No. 2, 299-341 (1990; Zbl 0697.60081)] derived asymptotic expansions to arbitrary order for the area of the Wiener sausage in terms of the self-intersection local times. In this paper the author shows that similar expansions are valid for the stable sausage. Let \(X_ t\) be a symmetric stable process in \(\mathbf R^ 2\) of order \(\beta,\;1<\beta<2\). The stable sausage of radius \(\epsilon >0\) is defined as \(S_ \epsilon (t) = \{ y\in\mathbf R :\inf_{0\leq s\leq t}| y-X_ s|\leq\epsilon\}\). Let \(\zeta\) be an exponential time of mean 1, let \(Q = P\otimes e^{-t} dt\) where \(P\) denotes the probability for the process starting at the origin, and let \(c_ 0(\epsilon) = c(\beta)\epsilon^{2-\beta}\) be the \(\beta\)-capacity of the disk of radius \(\epsilon\) centered at the origin. If \((2k-2)(2- \beta)<1\), there exist random variables \(\gamma_ n\in L^ 2(Q),\;n\leq k\), such that the following asymptotic expansion of the area of the sausage is valid: \[ m(S_ \epsilon(\zeta)) = \sum_{n=1}^ k (- 1)^{n-1} c_ 0^ n(\epsilon)\gamma_ n(\zeta) + o(c_ 0^ k(\epsilon)) \] in \(L^ 2(Q)\). Random variables \(\gamma_ n\) are renormalized self-intersection local times. In contrast to the case of the Wiener sausage, there is an upper bound to the order of expansion due to the fact the approximate renormalized self-intersection local times do not converge in \(L^ 2(Q)\) for larger \(k\) [see \textit{J. Rosen}, Stochastic processes, Proc. Semin., Vancouver/Can. 1990, Prog. Probab. 24, 285-320 (1991)]. The same expansion in \(L^ 2(P)\) (with slightly different condition on \(k\)) is valid when the random time \(\zeta\) is replaced by the fixed time \(t\). By explicitly computing the expectation of random variables \(\gamma_ n(t)\in L^ 2(P)\), the author obtains a precise asymptotic expansion of \(E(m(S_ 1(t)))\) as \(t\to\infty\).
0 references
sausage for stable processes
0 references
asymptotic expansions
0 references
Wiener sausage
0 references
self-intersection local times
0 references