Browse Prior Art Database

Association of Voice Activity Detection and Echo Suppress

IP.com Disclosure Number: IPCOM000043041D
Original Publication Date: 1984-Jul-01
Included in the Prior Art Database: 2005-Feb-04
Document File: 2 page(s) / 35K

Publishing Venue

IBM

Related People

Desblache, A: AUTHOR [+3]

Abstract

In a digital speech transmission system, the 2 to 4 wire conversion is performed by a transformer (hybrid). However due to the imperfect impedance matching (which depends on the calling line), the received signal is partially reflected and goes through the transmitting path, thus generating an echo. Echo control must therefore be implemented so as to avoid this inconvenience. Two basic methods may be used: Echo cancellers and Echo suppressors (ES). We address here the association of the latter with a Voice Activity Detector (VAD) function which is generally performed in TASI-based systems or, more generally, in a number of multiplexers. The principle of the proposed algorithm makes use of the well-known echo suppress method where the energy of the local speech (A1) is compared to the energy of the received signal (A2).

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 100% of the total text.

Page 1 of 2

Association of Voice Activity Detection and Echo Suppress

In a digital speech transmission system, the 2 to 4 wire conversion is performed by a transformer (hybrid). However due to the imperfect impedance matching (which depends on the calling line), the received signal is partially reflected and goes through the transmitting path, thus generating an echo. Echo control must therefore be implemented so as to avoid this inconvenience. Two basic methods may be used: Echo cancellers and Echo suppressors (ES). We address here the association of the latter with a Voice Activity Detector (VAD) function which is generally performed in TASI-based systems or, more generally, in a number of multiplexers. The principle of the proposed algorithm makes use of the well- known echo suppress method where the energy of the local speech (A1) is compared to the energy of the received signal (A2). If A1 < A2, then the transmitted signal is supposed to be an attenuated version (echo) of the received signal; thus, a hangover time N, equal to the delay t introduced within the analog path, is set up, (e.g., N = 3 x 20 ms blocks of samples), and the N following blocks are reset to 0. In the other case, the transmitted signal may be either a speech signal or a local background noise. This ambiguous situation should be resolved by the VAD before any action on the echo suppressor is taken.

1

Page 2 of 2

2

[This page contains 2 pictures or other non-text objects]