Signal understanding: an artificial intelligence approach to modulation classification

Author

Whelchel, J.E. ; McNeill, D.L. ; Hughes, R.D. ; Loos, M.M.

Author_Institution

E-Syst. Melpar Div., Falls Church, VA, USA

fYear

1989

fDate

23-25 Oct 1989

Firstpage

231

Lastpage

236

Abstract

A signal understanding system being developed and evaluated by means of simulation is described. The system consists of a new type of generalized demodulation/feature extraction section followed by a statistical moment generator, and then by a pattern classifier. The classifier is based on neural network topology. An algorithmic maximum-likelihood (ML) pattern classifier was also used to evaluate the performance of the neural network. Signal generation, feature extraction, and the ML classifier were implemented using VAX/VMS Fortran. The neural network was implemented on a PC-based Hecht-Neilson ANZA neurocomputer

Keywords

computerised pattern recognition; computerised signal processing; neural nets; ML classifier; PC-based Hecht-Neilson ANZA neurocomputer; VAX/VMS Fortran; algorithmic maximum-likelihood; artificial intelligence; feature extraction; modulation classification; network training; neural network topology; pattern classifier; signal understanding system; statistical moment generator; Artificial intelligence; Demodulation; Feature extraction; Filtering; Frequency estimation; Maximum likelihood estimation; Neural networks; Phase estimation; Quadrature phase shift keying; Surveillance;

fLanguage

English

Publisher

ieee

Conference_Titel

Tools for Artificial Intelligence, 1989. Architectures, Languages and Algorithms, IEEE International Workshop on

Conference_Location

Fairfax, VA

Print_ISBN

0-8186-1984-8

Type

conf

DOI

10.1109/TAI.1989.65325

Filename

65325