A characterization and moving average representation for stable harmonizable processes (Q1815750): Difference between revisions

Summary: We provide a characterization for symmetric \(\alpha\)-stable harmonizable processes for \(1< \alpha \leq 2\). We also deal with the problem of obtaining a moving average representation for stable harmonizable processes discussed by \textit{S. Cambanis} and the second author [Z. Wahrscheinlichkeitstheorie Verw. Geb. 66, 593-612 (1984; Zbl 0528.60035)], \textit{A. Makegon} and \textit{V. Mandrekar} [Probab. Theory Relat. Fields 85, No. 1, 1-11 (1990; Zbl 0673.60041)] and \textit{S. Cambanis} and \textit{C. Houdré} [ibid. 95, No. 1, 75-85 (1993; Zbl 0794.60027)]. More precisely, we prove that if \(Z\) is an independently scattered countable additive set function on the Borel field with values in a Banach space of jointly symmetric \(\alpha\)-stable random variables, \(1<\alpha\leq 2\), then there is a function \(k\in L^2 (\lambda)\) \((\lambda\) is the Lebesgue measure) and a certain symmetric \(\alpha\)-stable random measure \(Y\) for which \[ \int^\infty_{-\infty} e^{itx} dZ(x) = \int^\infty_{-\infty} k(t-s) dY(s), \quad t\in \mathbb{R}, \] if and only if \(Z(A)=0\) whenever \(\lambda (A)=0\). Our method is to view \(\text{S} \alpha \text{S}\) processes with parameter space \(R\) as \(\text{S} \alpha \text{S}\) processes whose parameter spaces are certain \(L^\beta\) spaces.