Optimal Sensor Scheduling for Multiple Hypothesis Testing
Abstract
The generic problem of selecting the sequence of sensors which optimizes the information received about a number of discrete hypotheses is considered. The optimization criterion penalizes the uncertainty present about pairs of hypotheses in a form which has an eigenfunction property with respect to a Bayes update of the conditional probability distribution. Application of the Portryagin minimum principle yields elegant solutions to an interesting class of problems. Applications in surveillance, failure detection, and nondestructive testing are possible.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1981
- Accession Number
- ADA113399
Entities
People
- Robert R. Tenney
Organizations
- Massachusetts Institute of Technology