A CODING CRITERION FOR CONTINUOUS CHANNELS DERIVED FROM LIKELIHOOD ESTIMATION

Abstract

The divergence between hypotheses is investigated as a criterion for signal selection, and is shown to be useful whenever the probability of error cannot be directly evaluated. After a brief discussion of the divergence in hypothesis-testing problems, the relation between the divergence and the information transfer of the channel is determined. As an illustration of the use of the divergence as a coding criterion, the criterion is applied to a finite set of known signals in additive noise. The divergence is evaluated for this example and compared with the probability of error in a maximum-likelihood receiver. (Author)

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1961
Accession Number
AD0257868

Entities

People

  • T.l. Grettenberg

Organizations

  • Lockheed Martin Missiles and Space

Tags

DTIC Thesaurus Topics

  • Additives (Chemicals)
  • Hypotheses
  • Information Processing
  • Information Transfer
  • Probability

Fields of Study

  • Engineering

Readers

  • Radar Systems Engineering.
  • Statistical inference.