For a nonnegative
measurable function f satisfying
define
Berman proved, extending so-called “Polya characteristic function”, that the r is
the characteristic function of an absolutely continuous distribution. The
positive-definiteness of the r corresponds to a stationary Gaussian process, which is
called Polya-Covariance process or simply Polya process.
In this paper, some analytic properties of its sample functions are studied: (1)
continuity, (2) differentiability, (3) quadratic variation, and (4) upper and lower
class.
|