A note on generalized Gaussian random fields (Q1108661)

From MaRDI portal
scientific article
Language Label Description Also known as
English
A note on generalized Gaussian random fields
scientific article

    Statements

    A note on generalized Gaussian random fields (English)
    0 references
    0 references
    1988
    0 references
    Let \({\mathcal D}\) be a bounded domain in \({\mathbb{R}}^ d\) with smooth boundary \(\partial {\mathcal D}\). The white noise on \({\mathcal D}\) is defined as a measure \(\mu\) on the Sobolev space \(H^{-m}({\mathcal D})\), \(m>d/2\), corresponding to a characteristic functional \[ C(\xi)=\exp [-2^{- 1}\int_{{\mathcal D}}\xi^ 2(t)dt],\quad \xi \in H^ m({\mathcal D}). \] The author discussed a generalized Gaussian random field \[ {\mathcal X}\equiv \{X(\zeta)=\int_{{\mathcal D}}x(t)\zeta (t)dt,\quad x\in H^{-m},\quad \zeta \in E\}, \] where E is a certain function space. Let e be a natural embedding from \(H^{m-1/2}(\partial {\mathcal D})\) to \(H^ m({\mathcal D})\), then Proposition: There exists a white noise measure \(\mu\) ' on \(H^{- m+1/2}(\partial {\mathcal D})\) which is linked with \(\mu\) as \((e^*)^{- 1}\cdot \mu =\mu '.\) Using this proposition, we can define a new class of random variables \(X(C)=\int_{C}x(t)G(t)dt\), \(C=\partial {\mathcal D}\), in connection with \({\mathcal X}\). The variation \(\partial X(C)\) attached with the d of these equations satisfy cohomological identities. A diffusion for which the coefficients differ in a cohomologically trivial sense from a given diffusion can be constructed by a perturbation procedure. Every quantum diffusion on the algebra of all bounded operators on a Hilbert space is characterized by a unitary process. In the commutative case certain ``diffusions'' with discontinuous sample paths are found; these are a feature of the zero temperature Fock quantum stochastic calculus used here and do not exist at finite temperature.
    0 references
    random field
    0 references
    characteristic functional
    0 references
    generalized Gaussian random field
    0 references
    natural embedding
    0 references
    Hilbert space is characterized by a unitary process
    0 references
    Fock quantum stochastic calculus
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references

    Identifiers