Likelihood Ratios for Sequential Decision-Making on Markov Sequences.
Abstract
The purpose of this research is to derive a variety of likelihood ratios for the detection of Markov sequences in noise. The program consists of exploiting the Bayesian recursions appropriate to a related filtering problem, together with a 'known-form' likelihood ratio, to obtain the desired result. In the derivation of a discrete-time Gauss-Markov likelihood ratio the authors seek a 'pure' causal estimator-correlator structure and encounter a 'locally stable' state estimator that is of some interest in its own right. The likelihood ratio is 'pure' in the sense that the locally stable estimator is used in precisely the same manner as the stored replica in a known-form detection problem. The estimator is 'locally stable' in the sense that it equalizes within a constant related to the a priori and a posteriori filtering error covariances, the prior and a posteriori filtering densities.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1975
- Accession Number
- ADA014667
Entities
People
- Loren W. Nolte
- Louis L. Louis L. Scharf
Organizations
- Colorado State University