Abstract
We consider a multi-channel system in which one apparatus makes a sequence of observations, one at a time. By means of scanning, i.e. selecting a channel to be analyzed at any instant and deciding to stop at some stage, it is required to determine the channel in which there is the signal with prescribed constraints on error probabilities. A simple scanning rule, based on a cyclic application of a sequential probability ratio test (SPRT) is proposed for this problem. It is proved that in the case of Brownian motion, the expected scanning time of this rule is equal to the one of the optimal scanning rule (which is known only for this case). The simple structure of this rule permits to obtain corrected Brownian approximations for its characteristics in the case of exponential family of distributions. The same procedure is used in multi-channel change point problem.
Similar content being viewed by others
References
Bassevile M, Nikiforov IV (1993) Detection of abrupt changes: Theory and applications. PTR Prentice-Hall New Jersey
Brodsky BE, Darkhovsky BS (1993) Nonparametric methods in change-point problems. Kluwer Academic Dordrecht
Dragalin V (1989) Asymptotic solution of a problem of signal searching in multi-channel system. Mat Issled (Academy of Sciences of Moldova) 109:15–35 (In Russian)
Dragalin V (1994) A multi-channel change point problem. Proceedings of 3rd Umeå-Würzburg Conference in Statistics. Umeå University
Lotov VI (1987) Asymptotic expansions in the sequential probability ratio test. Theory Probab App 32:62–72
Pollak M, Siegmund D (1985) A diffusion process and its application to detecting a change in the drift of Brownian motion. Biometrika 72:267–280
Shiryaev AN (1963) On the detection of disorder in a manufacturing process. Theory Probab Appl 8:247–265, 402–413
Shiryaev AN (1964) On the theory of decision function and control by observation from incomplete data. Trans Third Prague Conf Inform Theory Prague 657–687 (In Russian)
Siegmund D (1985) Sequential analysis: Tests and confidence intervals. Springer Berlin
Srivastava MS, Wu Y (1993) Comparison of EWMA, CUSUM and Shiryayev-Roberts procedures for detecting a shift in the mean. Ann Statist 21:645–670
Woodroofe M (1982) Nonlinear renewal theory in sequential analysis. SIAM Philadelphia
Zigangirov KSh (1966) On a problem in optimal scanning. Theory Probab Appl 11:294–298
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Dragalin, V. A simple and effective scanning rule for a multi-channel system. Metrika 43, 165–182 (1996). https://doi.org/10.1007/BF02613905
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02613905