SEQUENTIAL DETECTION,
Abstract
Basic problems involving the binary, sequential detection of signals in noise are discussed in detail; earlier theory is briefly reviewed and then extended to take into account broader types of models: in particular, those cases where the signal is unknown but the distribution of signal parameters is given or available. New units for the following specific sequenaal examples are given: (1), independent sampling, incoherent observation and a Rayleigh d.d. of amplitudes; (00, dependent, incoherent observation, with known signal amplitudes; (3), same as (2), but with a Rayleigh d.d. of amplitudes. Both dis crete and continuous sampling are employed. Comparisons are made with the corresponding fixed-sample cases, and it is observed (as ex pected) that these sequential procedures are, on the average, superior to the fixed-sample detectors (i.e. require smaller input signals, etc. for the same expected performance). The main attention is given here to the average sample numbers (ASN) and to the Operating Char acteristic Functions (OCF) of these sequential tests. OCF's for fixed-sample situations are considered in the one- and two-threshold cases, also. Some miscellaneous results for canonical threshold, sequential detectors are included, as well. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1963
- Accession Number
- AD0416724
Entities
People
- David Middleton
Organizations
- Johns Hopkins University