Title :
Detection of partial response signals using noise prediction techniques
Author :
D. Makrakis;A. Yongacoglu;K. Feher
Author_Institution :
Commun. Res. Centre, Ottawa, Ont., Canada
fDate :
6/1/1991 12:00:00 AM
Abstract :
New symbol-by-symbol detectors for partial response signalling systems are introduced. The proposed detectors are based on noise prediction techniques used to improve the detection efficiency. The purpose is to approach to the performance of the maximum likelihood detector with a less complex structure. This is achieved by designing receivers having one or two stages of detection which utilise symbol-spaced or fractionally-spaced samples. The principle of the new detectors is demonstrated for the duobinary system. The method is also applicable to other partial response signalling systems. Performance evaluation results indicate gains in the range of 1.2 dB for one-stage and 1.6 dB for two-stage detectors.
Keywords :
"Filtering","Signal detection","White noise","Prediction methods"
Journal_Title :
IEE Proceedings I - Communications, Speech and Vision
DOI :
10.1049/ip-i-2.1991.0017