Homomorphic detection of convolved signals using class 1 filters

It is shown that signals which are convolutionally combined can be detected using optimal homomorphic filtering. The system uses Fourier transforms combined with a log( ) operator. Dual-channel filtering is described. Single-channel filtering is also possible but does not perform as well. A simulati...

Full description

Saved in:
Bibliographic Details
Published inThirty-Fourth Midwest Symposium on Circuits and Systems, 1991 pp. 486 - 489 vol.1
Main Authors Lindquist, C.S., Cox, S.W.
Format Conference Proceeding
LanguageEnglish
Published IEEE 1991
Subjects
Additive noise
Biology computing
Convolution
Fourier transforms
Frequency domain analysis
Nonlinear filters
Phase detection
Signal detection
Speech processing
Wiener filter
Online AccessGet full text
ISBN9780780306202
0780306201
DOI10.1109/MWSCAS.1991.252201

Cover

More Information
Summary:It is shown that signals which are convolutionally combined can be detected using optimal homomorphic filtering. The system uses Fourier transforms combined with a log( ) operator. Dual-channel filtering is described. Single-channel filtering is also possible but does not perform as well. A simulation is presented to demonstrate the computations involved. Although Class 1 algorithms were used for simplicity, the techniques have been extended using Class 2 algorithms. Excellent results are also obtained. This homomorphic filtering technique can be used to solve a variety of problems that cannot otherwise be solved using classical methods.< >
ISBN:9780780306202
0780306201
DOI:10.1109/MWSCAS.1991.252201