Browse Prior Art Database

Self Determination of Error Rate in Data Receivers

IP.com Disclosure Number: IPCOM000052543D
Original Publication Date: 1981-Jun-01
Included in the Prior Art Database: 2005-Feb-11
Document File: 2 page(s) / 35K

Publishing Venue

IBM

Related People

Godard, D: AUTHOR

Abstract

The figure illustrates a method of self-determination of error rate in a data receiver incorporating a transversal equalizer with predictive error filtering, based on two independent estimations of transmitted symbols.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 54% of the total text.

Page 1 of 2

Self Determination of Error Rate in Data Receivers

The figure illustrates a method of self-determination of error rate in a data receiver incorporating a transversal equalizer with predictive error filtering, based on two independent estimations of transmitted symbols.

In today's medium and high-speed modems, transmission quality is determined by observing the eyemeter at the equalizer output, and can only be qualified as "good", "Marginal" or "bad". However, in the presence of impairments, such as impulse noise or third-order harmonic distortion, the eyemeter value is a very poor indicator of error probability The goal of the proposed method is to provide modems with the capability of self-determining the error rate with a high accuracy, and thus to help in network monitoring. The error rate can be displayed on the operator panel.

Decision-feedback equalization is known to offer better tolerance to linear distortions than transversal equalization. It can be shown theoretically that the performance of the decision-feedback equalizer can be attained by applying to the transversal equalizer the concept of error prediction.

Let z(n) be the complex equalizer output signal. It may be written as: (1) z(n) = a(n) + e(n) where a(n) is the transmitted data symbol and e(n) the error signal. Since in general e(n) is a colored random process, the previous errors e(n-k) may be used to diminish its mean-squared value: a new error signal Epsilon(n) is obtained through See origi...