Positivity and positive definiteness in generalized function algebras (Q864647)

Let \(\Omega\) be an open set in \(\mathbb{R}^n\) and \({\mathcal G}(\Omega)\) be the special Colombeau algebra. The authors give and analyse definitions for \(f\in{\mathcal G}(\Omega)\) to be positive or \({\mathcal D}'(\Omega)\)-weakly positive, and also for \(f\in{\mathcal G}(\mathbb{R}^n)\) to be positive definite. Positivity is defined as follows: \(f=[(f_\varepsilon)_\varepsilon]\in {\mathcal G}(\Omega)\) is said to be positive if there exists a representative \((f_\varepsilon)_\varepsilon\) of \(f\) such that for every \(a> 0\) and \(K\subset\subset\Omega\), there exists \(\varepsilon_0\in (0,1)\) such that \(\text{inf\,}f_\varepsilon(x)+ \varepsilon^a\geq 0\), \(x\in K\), \(\varepsilon< \varepsilon_0\). The authors also consider the same notions in the space of Colombeau tempered generalized functions. As an example of possible applications, they prove a Bochner-type theorem and pose the following open problem: Does there exist \(G= [(G_\varepsilon)_\varepsilon]\in{\mathcal G}_\tau(\mathbb{R})\) that is tempered weakly positive definite but so that \(F= [({\mathcal F}^{-1}(G_\varepsilon)*\phi_\varepsilon)_\varepsilon]\) is not tempered weakly positive? The paper is written concisely and is rich in new and interesting results.