ITU-T G 167-1993 Acoustic Echo Controllers (Study Group XV 19 pp)《声回波控制器(研究15组)19pp》.pdf

1、ITU-T RECflN*G*367 93 = 4862591 0582236 7T7 W INTERNATIONAL TELECOMM U N CATION UN ION ITU-T TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU G.167 (03/93) GENERAL CHARACTERISTICS OF INTERNATIONAL TELEPHONE CONNECTIONS AND INTERNATIONAL TELEPHONE CIRCUITS ACOUSTIC ECHO CONTROLLERS ITU-T Recommendatio

2、n G.167 (Previously “CCITT Recommendation”) - ITU-T RECMN*G.367 73 H 4862573 0582237 633 I FOREWORD The ITL Telecommunication Standardization Sector (ITIJ-T) is a permanent organ of the. International Telecom- inunication Union. The ITU-T is responsible for studying technical, operating and tariff q

3、uestions and issuing Recommendations on them with a view to standardizing telecommunications on a worldwide basis. The World Teleconimunication Standardization Conference (WTSC), which meets every four years, established the topics for study by the ITU-T tudy Groups which, in their turn, produce Rec

4、ommendations on these topics. ITU-T Recommendation G.167 was prepared by the ITGT-T Study Group XV (1988-1993) and was approved by the WTSC (Ilelsinlu, March 1-12, 1993). NOTES 1 As a consequence of a reform process within the International Telecommunication Union (ITU), the CCITT ceased to exist as

5、 of 28 February 1993. In its place, the ITU Telecommunication Standardization Sector (ITU-T) was created as of 1 March 1993. Similarly, in this reform process, the CCIR and the IFRB have been replaced by the Radiocoinmunication Sector. In order not to delay publication of this Recommendation, no cha

6、nge has been made in the text to references containing the acronyms “CCITT, CCIR or IFRB” or their associated entities such as Plenary Assembly, Secretariat, etc. Future etlirions of this Recommendation will contain the proper terminology related to the new ITU structure. 2 telecommunication adminis

7、tration and a recognized operating agency. In this Recommendation, the expression “Administration” is used for conciseness to indicate both a G ITU 1993 All rights reserved. No pait of this publication may be reproduced or utilized in any form or by any means. electronic or mechanical, including pho

8、tocopying and microfilm, without permission in writing from the ITU. ITU-T RECNN*G.167 93 Y8b259L 0582218 57T = CONTENTS General . 1.1 Definition of acoustic echo controllers . 1.2 Applicability of the Recommendation 1.3 Objective of the Recommendation . 1.4 Relevant Recommendations . General defini

9、tions . 2.1 Audio terminals 2.2 Interfaces 2.3 Processing unit 2.3 Electroacoustic transducers and attached circuits . Definitions relating to acoustic echo controllers 3.2 Items relevant to speech performance of acoustic echo controllers Transmission specifications . 4.1 scope 4.2 Application speci

10、fics 4.3 Measurements . 4.4 Bandwidth . 4.5 Attenuation distortion . 4.6 Delay . 1.7 Delay distortion 4.8 Non-linear distortion . 4.9 4.10 Acoustic echo control specifications 5.1 Scope 5.2 Measurement conditions . 5.3 5.4 Specifications for interworking with the network 6.1 Scope 6.2 6.3 Interworki

11、ng with network echo cancellers 6.4 DCME and PCME interactions . 6.5 Noise emitted by the AEC processing unit on the send side Acoustic noise produced by the AEC on the receive side Correspondence between performance values and transmission delays . Specifications and verification tests Intenvorking

12、 with speech codecs Interworking between a wide-band terminal and other types of terminals through the network . Page 1 1 1 1 1 2 2 2 2 3 3 4 5 5 6 6 6 6 6 7 7 7 7 7 7 7 9 9 14 14 14 15 16 16 Recommendation 6.167 (03193) i ITU-T RECMN*GmLb7 93 m 48b2591 0582239 406 m Recommendation 6.167 ACOUSTIC EC

13、HO CONTROLLERS1) (Helsinki, 1993) 1 General 1.1 acoustic echo controllers (denoted by AECs) are voice operated devices installed in audio terminals on the customer premises, used for the purpose of eliminating acoustic echoes and protecting the communication from howling due to acoustic feedback fro

14、m loudspeaker to microphone. Definition of acoustic echo controllers 1.2 Applicability of the Recommendation This Recommendation is applicable to the design of AECs for audio terminals with digital or analogue line interfaces, and is intended for use in the following areas of telecommunications (den

15、oted by applications): - teleconferencing; - loudspeaking (hands-free) telephones; - videophone terminals; - mobile and personal applications. 1.3 Objective of the Recommendation The Recommendation specifies performance characteristics and values with which AECs must comply, and methods to verify th

16、ese performances. The performances depend on the applications considered. Particular processing techniques are indicated as guidance for possible iniplementations, but are not compulsory. 1.4 Relevant Recommendations The following Recommendations are relevant to the problem of acoustic echo control

17、in audio terminals: - Rec. P.30: Transmission performance of group audio terminals (GATS) - Rec. P.3 1: Transmission characteristics of digital telephones - Rec. P.34: Transniission characteristics of hands-free telephones - - Rec. G.131: Echo and delay Rec. G.173: Transmission planning aspects of t

18、he speech service in digital public land mobile networks. In case of digital transmission. the standard digital speech coding formats considered are: G.7 11 (telephone band), G.722 (wide band). For mobile applications, the coding law is under study. 1 J aiuei in brackets 1 are provisional. Recommend

19、ation 6.167 (03/93) 1 ITU-T RECNN*G.167 93 B 4862591 0582220 128 2 General definitions 2.1 Audio terminals Audio terminals are designed for hands-free voice communication between individuals or groups of persons through analogue or digital networks. Any audio terminal is functionally accessed throug

20、h user interfaces, network interfaces and test interfaces. The processing unit, electroacoustic transducers and attached circuits are internal parts of the terminal. 2.2 Interfaces The interfaces are places (outside the terminal) or points (inside the terminal) where physical measurements can be don

21、e in order to adjust internal parameters of the terminal for proper operation and to verify performance. The items measured at the interfaces are used to infer the subjective quality that the local user(s) and the distant user(s) would experience. 2.2.1 User interfaces There are two user interfaces:

22、 - Receive interface (Rout): The place(s) where acoustic attributes relating to characteristics of speech listened to by the local user(s) are measured. This is also referred to as the position of the measuring microphone as described in Recommendation P.34. Send interface (Sin): The place(s) where

23、acoustic attributes relating to characteristics of speech produced by the local user(s) are measured. This is referred to as the mouth reference point (MRP) in Recommendation P.34. - 2.2.2 Network interfaces There are two nehvork interfaces: - - Receive interface (Rin): A point where the electrical

24、signals received from the network are available. Send interface (S,J: A point where the electrical signals sent to the network are available. If the terminal is connected to an analogue line, the interfaces shall comply with the characteristics specified in Recommendation Q.552. If the terminal is c

25、onnected to a digital line (ISDN), the interfaces shall comply with the S-intelface characteristics specified in Recommendations 4.554 and G.7032). 2.2.3 Test interfaces The test interfaces are input and output points inside the terminal where signals and/or controls can be applied or measured for t

26、he purpose of performance verification. 2.3 Processing unit The processing unit includes all the devices in the terminal performing signal processing functions on audio signals (except the devices which are parts of the electroacoustic transducers and attached circuits). A partial list of functions

27、which may be in the processing unit is: - - - - AD and D/A linear conversions of audio signals; signal processing for acoustic echo control; signal processing for other purposes (e.g. noise cancellation, room reverberation reduction); transcoding of audio signals between line code format and linear

28、code. 2, In case of analogue line, the measuring equipment used for performance verification on the line side shall provide signal separation of both transmission directions by at least 60 dB for all frequencies in the transmission bandwidth (e.g. by using an adaptive hybrid in it). In case of digit

29、al line, this measuring equipment shall implement the same speech coding law as the one used in the terminal itself (e.g. Recommendations G.711, G.722, etc.) in the purpose to perform measurements on linear signals. 2 Recommendation 6.167 (03/93) ITU-T RECMN*G-167 93 W 4862593 0582221 Ob4 W 2.4 Elec

30、troacoustic transducers and attached circuits The electroacoustic transducers arc the loudspeaker(s) and niicrophone(s) attached to the terminal in nornial operation. The attached circuits may include the amplifiers, switches, level adjustments and other devices which can be under the control of the

31、 user or automatically adjusted during the operation of the terminal like sound equalizers, etc. The functional block diagram of a general audio terniinal equipped with an acoustic echo controller is shown in Figure 1. fn =I 1 .- .- e o O 2 :I m Processing unit Network S0“t Contrds FIGURE UG.167 Gen

32、eral audio terminal with AEC 3 Definitions relating to acoustic echo controllers For the purpose of this Reconiiiieiidation, the following definitions apply: 3.1 functional units: The functional units of an AEC are devices or parts of devices implemented in the processing unit, which contribute to t

33、he general function of acoustic echo control. There is no restriction on how to implement theni. The following subclauses describe functional units which can be parts of an AEC (not a complete list). 3.1.1 and distmt users speech. It is generally iiiipleniented by adaptive identification of the acou

34、stic echo path response. acoustic echo canceller: A device which reduces the acoustic echo level with negligible effects on the local 3.1.2 and/or transmitted audio signals. loss controller: A device which reduces the acoustic echo level by inserting variable losses on the received 3.1.3 haiiiples o

35、f the transmitted audio signal. A center-clipper is a typical device of this kind. non-linear processor: A device which reduces or cancels sniall echo signals by non-linear operation on the 3.1.4 supplementary howling control device: A device which modifies sonie characteristics of the transmitted a

36、nd/or received signals in order to improve the srability inargin of the terminal. This function is typically inipleniented by an harmonic processor. To prevent network disturbances, such devices should be avoided in the case of terminals likely to be used on connections including network electric ec

37、ho cancellers conforming to Reconmendation G.165 which are not able to work properly with tinie-variant echo paths. Recommendation 6.167 (03/93) 3 ITU-T RECMN*G.L67 93 B 4862.591 0582222 TTO a8a 3.1.5 Summary The functional units defined above can be combined for better performance. They can use all

38、 the available signals in the terminal (for example, the individual signals coming from several microphones arranged in an acoustic array). Moreover, they can be mixed with other functions (for example sub-band speech coding) for efficient implementation, provided that they do not modify the proper

39、characteristics of these functions when they are operating. It must be possible to enable, disable and reset functionally those devices when required by the test procedures described below in the Recommendation. A functional block-diagram of a typical processing unit is shown in Figure 2. L r bP To

40、network Howling control Non-linear interface From b Variableloss device processor si, -+ Acoustic echo estimator and other control circuits -r T Loss control ler I Variable loss To ?bo II I I A I T1507370-92/dO2 Y FIGURE 2/G.167 Functional block-diagram of a typical processing unit (AEC part) (bp de

41、notes bypass signal paths for testing purposes) 3.2 The following subclauses are definitions of items which are specifically linked to the static and dynamic performance of AECs. Values and test procedures to measure these items are given in 5. 3.2.1 weighted terminal coupling loss - single talk (TC

42、Lwst): The weighted loss between the Ri, and Saut network interfaces when the AEC is in normal operation, and when there is no signal coming from the local user3). 3.2.2 weighted terminal coupling loss - double talk (TCLwdt): The weighted loss between the Ri, and Saut network interfaces when the AEC

43、 is in normal operation, and where the local user and the far-end user are active simiiltaneously3). Items relevant to speech performance of acoustic echo controllers 3, The weighting is made according to the rule specified in Recommendation G.122 (computation of talker echo loudness rating). Care m

44、ust be taken to avoid possible masking of singing effects by the weighting (under study). 4 Recommendation G.167 (03/93) ITU-T RECNN*G-Lb? 93 W 48b259L 0582223 937 3.2.3 which i inserted by the AEC during double-talk events. The frequency response on the receive side during double talk is under stud

45、y. 3.2.4 inwrted by the AEC during double-talk events. The frequency response on the send side during double talk is under study. 3.2.5 received speech distortion during double talk (Drdt): The total non-linear signal distortion at the point which can be produced by the AEC during double-talk events

46、. 3.2.6 sent speech distortion during double talk (Dsdt): The total non-linear signal distortion at the Saut point which can be produced by the AEC during double-talk events. 3.2.7 frequency shifting (or pitch ratio Pr): The shift-up of frequencies of the signal at the points, due to howling control

47、 devices like harmonic processors niore details in 5.4.7). 3.2.8 break-in time - simple talk (Tonst): The tim interval between the onset of the received signal (similarly the traminitted signal) and the instant when the attenuation on the receive path (similarly on the send path) reaches 3 dB. Rir t

48、hi purpose, the other side is quiet. 3.2.9 break-in time - double talk (Tondt): The time interval between the onset of the received signal (siniilarly the ent signal) and the instant when the attenuation on the receive path (similarly on the send path) reaches the value Ardt similarly Ad). For this

49、purpose, the signal in the opposite direction of transmission is held at a specified level. 3.2.10 initial convergence time (Tic): The time interval between the instant when a specified test signal is applied to the R, port of the terniinal (after all the functions of the AEC have been reset and then enabled), and the instant when the returned echo signal at the S, port is attenuated by at least a pre-defined amount. The local user is not active. 3.2.11 recovery time after double talk (Trdt): The time elapsed between the end of a double talk event and the intmt when the