Optimal detection of a signal with unknown appearance and disappearance times (Q5951360)

The authors consider the problem of detecting a determinate signal with unknown (or random) appearance and disappearance times in Gaussian white noise for countinuous-time observation. They present maximum likelihood and optimal (Bayesian) algorithms for detection and estimation of appearance and disappearance times. They study asymptotic properties of the maximum likelihood algorithm and its efficiency. They obtain asymptotic expressions for characteristics of the maximum likelihood algorithms. By means of statistical computer modelling, they investigate efficiency of both algorithms as well as determine the applicability limits of the asymptotic expressions for characteristics of the algorithms.