arXiv Analytics

Sign in

arXiv:math/0609682 [math.PR]AbstractReferencesReviewsResources

On the second moment of the number of crossings by a stationary Gaussian process

Marie F. Kratz, José R. León

Published 2006-09-25Version 1

Cram\'{e}r and Leadbetter introduced in 1967 the sufficient condition \[\frac{r''(s)-r''(0)}{s}\in L^1([0,\delta],dx),\qquad \delta>0,\] to have a finite variance of the number of zeros of a centered stationary Gaussian process with twice differentiable covariance function $r$. This condition is known as the Geman condition, since Geman proved in 1972 that it was also a necessary condition. Up to now no such criterion was known for counts of crossings of a level other than the mean. This paper shows that the Geman condition is still sufficient and necessary to have a finite variance of the number of any fixed level crossings. For the generalization to the number of a curve crossings, a condition on the curve has to be added to the Geman condition.

Comments: Published at http://dx.doi.org/10.1214/009117906000000142 in the Annals of Probability (http://www.imstat.org/aop/) by the Institute of Mathematical Statistics (http://www.imstat.org)
Journal: Annals of Probability 2006, Vol. 34, No. 4, 1601-1607
Categories: math.PR
Subjects: 60G15, 60G10, 60G70
Related articles: Most relevant | Search more
arXiv:2004.12290 [math.PR] (Published 2020-04-26)
Sojourns of Stationary Gaussian Processes over a Random Interval
arXiv:1709.06760 [math.PR] (Published 2017-09-20)
Exponential concentration for zeroes of stationary Gaussian processes
arXiv:1011.6355 [math.PR] (Published 2010-11-29)
Exact asymptotics of supremum of a stationary Gaussian process over a random interval