声回波抵消技术的研究
|
摘要
随着通讯、数字信号处理和大规模集成电路技术的飞速发展，人们对声音通讯系统(比如视频会议系统等)中的话音质量提出了越来越高的要求。然而，使这些系统能达到令人满意的话音质量并不是一件容易的事情。由于存在从扬声器到麦克风的声音反馈路径，为了得到满意的听音水平，最重要的需求就是应使这些系统具有抗自激和回波信号的能力。因此，回波消除技术也就成为世界各大通讯公司竞争的热点技术之一。
本文主要研究用于声音通讯系统中的声回波抵消技术。声回波抵消通常采用声回波抵消器来实现，最简单的声回波抵消器由自适应滤波器组成。我们可以借助于它宋估计回波信号，并从麦克风信号中减掉该估计值来实现声回波的抵消。
全文共分六章：
第一章简要介绍了有关声回波抵消器的背景知识，包括声回波产生机理，声回波抵消原理及声回波抵消器的基本构成模块。最后，介绍了本论文所做的主要工作。
第二章主要讨论了各种自适应滤波算法。分析了最小均方误差(LMS)算法、归一化的最小均方误差(NLMS)算法和递归的最小二乘(RLS)算法，并且结合声回波抵消算法评估了它们的性能。
为了进一步降低声回波抵消算法的计算复杂性和改善其收敛特性，在第三章中，讨论了各种子带滤波算法，比如DFT子带和FIR子带等。最后，将归一化最小均方误差(NLMS)算法和子带技术相结合，介绍了子带归一化最小均方误差(Subband NLMS)算法。
第四章主要讨论了声回波抵消系统中的各种实现问题，比如双讲检测、有限精度的影响、非线性处理和自激信号的检测与控制。
最后两章给出了具体的实验结果和未来工作展望。其中主要测试指标为回波消去量(ERLE)和收敛速度。
With the rapid development of communication techniques, digital signal processing and VLSI techniques, people demand higher and higher speech quality in communication. However, it is not always easy to implement acoustic equipments for communication system with satisfactory speech quality. The most important requirement is to protect against howling and echo, due to acoustic feedback from loudspeaker to microphone, and to ensure a sufficient listening level. So acoustic echo cancellation (AEC) technique has become a hot issue of competition in the famous communication company all over the world.
This paper is focused on the AEC in communication system, e.g. videoconference system. Acoustic echo cancellation is normally achieved by means of an acoustic echo canceller, which, in its simplest form, consists of an adaptive filter which estimates the echo signal and subtracts this estimate from the microphone signal.
This paper is organized as follows:
The background knowledge about AEC such as acoustic echo mechanism, principle of AEC and basic components of AEC is introduced in Chapter 1.
In Chapter 2, many kinds of adaptive filter algorithms are discussed. This section analyzes the Least Mean Squares (LMS), Normalized LMS (NLMS) and Recursive Least Squares (RLS) algorithms, and evaluates their performance as acoustic echo canceller.
In order to further reduce the computation complexity and improve the convergence behavior, many sorts of subband filter algorithms are depicted in Chapter 3. Finally, associated with NLMS algorithm, the suband NLMS (SNLMS) algorithm is proposed.
In Chapter 4, various implementation issues associated with an acoustic echo cancellation system, such as double-talk detection, finite precision effects, nonlinear processing and howling detection and control are discussed.
Finally, experiment results are given to demonstrate the performance of the proposed AEC algorithm, where ERLE and rate of convergence are the main indexes.
本文主要研究用于声音通讯系统中的声回波抵消技术。声回波抵消通常采用声回波抵消器来实现，最简单的声回波抵消器由自适应滤波器组成。我们可以借助于它宋估计回波信号，并从麦克风信号中减掉该估计值来实现声回波的抵消。
全文共分六章：
第一章简要介绍了有关声回波抵消器的背景知识，包括声回波产生机理，声回波抵消原理及声回波抵消器的基本构成模块。最后，介绍了本论文所做的主要工作。
第二章主要讨论了各种自适应滤波算法。分析了最小均方误差(LMS)算法、归一化的最小均方误差(NLMS)算法和递归的最小二乘(RLS)算法，并且结合声回波抵消算法评估了它们的性能。
为了进一步降低声回波抵消算法的计算复杂性和改善其收敛特性，在第三章中，讨论了各种子带滤波算法，比如DFT子带和FIR子带等。最后，将归一化最小均方误差(NLMS)算法和子带技术相结合，介绍了子带归一化最小均方误差(Subband NLMS)算法。
第四章主要讨论了声回波抵消系统中的各种实现问题，比如双讲检测、有限精度的影响、非线性处理和自激信号的检测与控制。
最后两章给出了具体的实验结果和未来工作展望。其中主要测试指标为回波消去量(ERLE)和收敛速度。
With the rapid development of communication techniques, digital signal processing and VLSI techniques, people demand higher and higher speech quality in communication. However, it is not always easy to implement acoustic equipments for communication system with satisfactory speech quality. The most important requirement is to protect against howling and echo, due to acoustic feedback from loudspeaker to microphone, and to ensure a sufficient listening level. So acoustic echo cancellation (AEC) technique has become a hot issue of competition in the famous communication company all over the world.
This paper is focused on the AEC in communication system, e.g. videoconference system. Acoustic echo cancellation is normally achieved by means of an acoustic echo canceller, which, in its simplest form, consists of an adaptive filter which estimates the echo signal and subtracts this estimate from the microphone signal.
This paper is organized as follows:
The background knowledge about AEC such as acoustic echo mechanism, principle of AEC and basic components of AEC is introduced in Chapter 1.
In Chapter 2, many kinds of adaptive filter algorithms are discussed. This section analyzes the Least Mean Squares (LMS), Normalized LMS (NLMS) and Recursive Least Squares (RLS) algorithms, and evaluates their performance as acoustic echo canceller.
In order to further reduce the computation complexity and improve the convergence behavior, many sorts of subband filter algorithms are depicted in Chapter 3. Finally, associated with NLMS algorithm, the suband NLMS (SNLMS) algorithm is proposed.
In Chapter 4, various implementation issues associated with an acoustic echo cancellation system, such as double-talk detection, finite precision effects, nonlinear processing and howling detection and control are discussed.
Finally, experiment results are given to demonstrate the performance of the proposed AEC algorithm, where ERLE and rate of convergence are the main indexes.
引文
[1] 殷福亮，刘恩彩等，回波抵消器研制报告，1997
[2] 殷福亮，宋爱军，数字信号处理C语言程序集，辽宁科学技术出版社，沈阳，1997
[3] [美]B．维德罗，S.D.史蒂诶斯著，王永德，龙宪惠译，自适应信号处理，四川大学出版社，1985
[4] 张贤达，现代信号处理，清华大学出版社，1995年5月
[5] ITU-T Recommendation G.167, Acoustic Echo Controllers, 1993
[6] ITU-T Recommendation G. 165, Echo Cancellers, 1993
[7] R.E. Crochiere and L. R. Rabiner, Multirate Digital Signal Processing, Prentice-Hall, 1983
[8] A.V. Oppenheim and R. W. Schafer, Digital Signal Processing, Prentice-Hall, 1975
[9] A.V. Oppenheim and R. W. Schafer, Digital Signal Processing, Prentice-Hall, 1975, (中译本，科学出版社，1978)
[10] S. Haykin, Adaptive Filter Theory, Prentice-Hall, 1993
[11] B. Widrow and S. D. Stearns, Adaptive Signal Processing, Prentice-Hall, 1985
[12] James K. Hardy, et al., High frequency circuit design, Reston Pub. Co., 1979
[13] N. B. Jones and J. D. Watson, Digital signal processing: principles, devices, and applications, P. Peregrinus Ltd., 1990
[14] S.M. Kou and J. Chen, New adaptive llR notch filter and its application to howling control in speakerphone system, Electronics Letters, April 1992, vol. 28, no. 8, pp. 764-766
[15] B.Widrow, et al., Adaptive noise canceling: principles and applications, Proc. IEEE, December 1975, pp. 1692-1316
[16] A.H. Sayed and T. Kailath, A state-space approach to adaptive RLS filtering, IEEE Signal Processing Magzine, 1994, vol. 11, pp. 18-60
[17] C. Carlemalm, F. Gustafsson and B. Wahlberg, On the problem of detection and discrimination of double talk and change in the echo path, ICASSP96, pp. 2742-2745
[18] G. K. Boray and M. D. Srinath, Conjugate gradient techniques for adaptive echo cancellation, ICASSP92, IV-13-16
[19] G. K. Boray and M. D. Srinath, Conjugate gradient techniques for adaptive filtering, IEEE Trans on Circuits and Systems-I: Fundamental Theory and Applications, Jan. 1992, vol. 39,
pp. 1-10
[20] G. Keratiotis and L. Lind, Optimum variable step-size sequence for LMS adaptive filters, IEE Proc-Vision, Image and Signal Processing, 1999, vol. 146, no.l
[21] H. Yasukawa, S. Shimada and I. Furukawa, Acoustic echo canceller with high speech quality, ICASSP87, pp. 49． 8． 1-49． 8． 4
[22] H. Yasukawa, Acoustic echo canceller with subband noise cancelling, Electronics Letters, July 1992, vol. 28, no. 15, pp. 1403-1404
[23] J. Benesty and P. Duhamel, A fast exact least mean square adaptive algorithm, IEEE Trans on Signal Processing, 1992, vol. 40, no. 12, pp. 2904
[24] J. R. Casar-Corredera and J. Alcazar-Fernandez, An acoustic echo canceller for teleconference systems, ICASSP86, pp. 1317-1329
[25] P. Heikamper, Optimization of an acoustic echo canceller combined with adaptive gain control, ICASSP95, pp. 3047-3050
[26] S. B. Gelfand, Y. Wei and J. V. Krogmeier, Stability of variable and random stepsize LMS, ICASSP97, pp. 1961-1964
[27] Shigenobu, A stereophonic echo canceller using single adaptive filter, ICASSP95, pp. 3027-3030
[28] S. L. Gay and R. J. Mammone, Fast converging subband acoustic echo cancellation using RAP on the WE DSP 16A, ICASSP90, pp. 1141-1144
[29] T. Petition, A. Gilloire and S. Theodoridis, Complexity reduction in fast RLS transversal adaptive filters with application to acoustic echo cancellation, ICASSP92, pp. Ⅳ-37-Ⅳ-40
[30] D. T. M. Slock, On the convergence behavior of the LMS and the normalized LMS algorithm, IEEE Transactions on Signal Processing, September 1993, vol. 41, no. 9, pp. 2811-2825
[31] M. Tarrab and A. Feuer, Convergence and performance analysis of the normalized LMS algorithm with uncorrelated Gaussian data, IEEE Transactions on Information Theory, July 1988, vol. 34, no. 4, pp. 680-691
[32] J. Prado and E. Moulines, Frequency-domain adaptive filter with application to acoustic echo cancellation, Annals de Telecommunications, 1994, vol. 49, no. 7-8, pp. 414-428
[33] E. Eleftheriou and D. D. Falconer, Tracking properties and steady-state performance of
RLS adaptive filter algorithms,IEEE Transactions on Acoustics,Speech,and Signal Processing,October 1986,vol.ASSP-34,no.5,pp.1097-1110
[34] D.R.Morgan and J.C.Thi,A delayless subband adaptive filter architecture,IEEE Transactions on Signal Processing,August 1995,vol.43,no.8,pp.1819-1830
[35] A.Gilloire,Experiments with subband acoustic echo cancelers for teleconferencing, Proceedings of ICASSP,Dallas,April 1987,pp.2141-2144
[36] K.Ashihara,K.Nishikawa and H.Kiya,Improvement of convergence speed for subband adaptive digital filters using multirate repeating method,Proceedings of ICASSP,Detroit, May 1995,pp.989-992
[37] R.Martin and J.Altenhoner,Coupled adaptive filters for acoustic echo control and noise reduction,Proceedings of ICASSP,Detroit,May 1995,pp.3043-3046
[38] V.A.Margo,D.M.Etter and J.H.Gross,Multiple short length adaptive filters for time varying echo cancellation,ICASSP 93,pp.1-161-164
[39] R.Storn,Echo cancellation techniques for multimedia applications-a Survey,TR-96-046, 1996
[40] J.D.Johnston,A filter family designed for use in quadrature mirror filter banks, ICASSP80,pp.291-294
[41] H.Nussbaumer and M.Veterli,Computaionally efficient QMF filter banks,ICASSP84,pp. 11． 3． 1-11． 3． 4
[42] W.Kellermann,Analysis and design of multirate systems for cancellation of acoustical echos,ICASSP88,pp.2570-2573
[43] F.A.Reed,P.L.Feintuch,A comparison of LMS adaptive cancellers implemented in the frequency domain and the time domain,IEEE Trans,on Circuits and Systems,June 1981, vol.CAS-28,no.6
[44] M.R.Asharif,T.Hayashi and K.Yamashita,Correlation LMS algorithm and its application to double-talk echo canceling,Electronics Letters,Feb.1999,vol.18,no.5,pp. 1062-1068
[45] K.Fujii and J.Ohga,A fast adaptive algorithm suitable for acoustic echo canceller,IEICE Trans.Fundamentals,Nov.1992,vol.E75-A
[46] T.Ganseler,A double talk resistant subband echo canceller,Signal Processing,1998, vol.65,pp.89-101
[47] T. Ganseler, M. Hansson, C. Ivarsson and G. Salomonsson, A double-talk detector based on coherence, IEEE Trans. on Communications, Nov. 1996, vol. 44 no. 11
[48] 藤井健作、大贺寿郎，音响有用推定误差安定的确保方法，电子情报通信学会论文志，EA-97，Nov．1997，pp．53-60
[49] H. Ye and B. Wu, A new double-talk detection algorithm based on the orthogonality theorem, IEEE Trans. on Communications, Nov. 1991, vol. 39, no. 11
[50] N. Nilsson, An echo canceller with frequency dependent NLP attenuation, Ericsson Radio Systems AB, Kista Sweden, Jun. 1998
[51] J.R. Deller, J. G. Proakis and J. H. L. Hansen, Discrete-time processing of speech signals, Macmillan Publishing Company, New York, 1993
[52] 宗孔德，多抽样率信号处理，清华大学出版社，1996
[53] 秦前清等，实用小波分析，西安电子科技大学出版社，1993
[54] B. Ayad, G. Faucon and R. L. Bouquin-Jeannes, Optimization of a noise reduction preprocessing in an acoustic echo and noise controller, Proceedings of ICASSP, Atlanta, May 1996, pp. 953-956
[2] 殷福亮，宋爱军，数字信号处理C语言程序集，辽宁科学技术出版社，沈阳，1997
[3] [美]B．维德罗，S.D.史蒂诶斯著，王永德，龙宪惠译，自适应信号处理，四川大学出版社，1985
[4] 张贤达，现代信号处理，清华大学出版社，1995年5月
[5] ITU-T Recommendation G.167, Acoustic Echo Controllers, 1993
[6] ITU-T Recommendation G. 165, Echo Cancellers, 1993
[7] R.E. Crochiere and L. R. Rabiner, Multirate Digital Signal Processing, Prentice-Hall, 1983
[8] A.V. Oppenheim and R. W. Schafer, Digital Signal Processing, Prentice-Hall, 1975
[9] A.V. Oppenheim and R. W. Schafer, Digital Signal Processing, Prentice-Hall, 1975, (中译本，科学出版社，1978)
[10] S. Haykin, Adaptive Filter Theory, Prentice-Hall, 1993
[11] B. Widrow and S. D. Stearns, Adaptive Signal Processing, Prentice-Hall, 1985
[12] James K. Hardy, et al., High frequency circuit design, Reston Pub. Co., 1979
[13] N. B. Jones and J. D. Watson, Digital signal processing: principles, devices, and applications, P. Peregrinus Ltd., 1990
[14] S.M. Kou and J. Chen, New adaptive llR notch filter and its application to howling control in speakerphone system, Electronics Letters, April 1992, vol. 28, no. 8, pp. 764-766
[15] B.Widrow, et al., Adaptive noise canceling: principles and applications, Proc. IEEE, December 1975, pp. 1692-1316
[16] A.H. Sayed and T. Kailath, A state-space approach to adaptive RLS filtering, IEEE Signal Processing Magzine, 1994, vol. 11, pp. 18-60
[17] C. Carlemalm, F. Gustafsson and B. Wahlberg, On the problem of detection and discrimination of double talk and change in the echo path, ICASSP96, pp. 2742-2745
[18] G. K. Boray and M. D. Srinath, Conjugate gradient techniques for adaptive echo cancellation, ICASSP92, IV-13-16
[19] G. K. Boray and M. D. Srinath, Conjugate gradient techniques for adaptive filtering, IEEE Trans on Circuits and Systems-I: Fundamental Theory and Applications, Jan. 1992, vol. 39,
pp. 1-10
[20] G. Keratiotis and L. Lind, Optimum variable step-size sequence for LMS adaptive filters, IEE Proc-Vision, Image and Signal Processing, 1999, vol. 146, no.l
[21] H. Yasukawa, S. Shimada and I. Furukawa, Acoustic echo canceller with high speech quality, ICASSP87, pp. 49． 8． 1-49． 8． 4
[22] H. Yasukawa, Acoustic echo canceller with subband noise cancelling, Electronics Letters, July 1992, vol. 28, no. 15, pp. 1403-1404
[23] J. Benesty and P. Duhamel, A fast exact least mean square adaptive algorithm, IEEE Trans on Signal Processing, 1992, vol. 40, no. 12, pp. 2904
[24] J. R. Casar-Corredera and J. Alcazar-Fernandez, An acoustic echo canceller for teleconference systems, ICASSP86, pp. 1317-1329
[25] P. Heikamper, Optimization of an acoustic echo canceller combined with adaptive gain control, ICASSP95, pp. 3047-3050
[26] S. B. Gelfand, Y. Wei and J. V. Krogmeier, Stability of variable and random stepsize LMS, ICASSP97, pp. 1961-1964
[27] Shigenobu, A stereophonic echo canceller using single adaptive filter, ICASSP95, pp. 3027-3030
[28] S. L. Gay and R. J. Mammone, Fast converging subband acoustic echo cancellation using RAP on the WE DSP 16A, ICASSP90, pp. 1141-1144
[29] T. Petition, A. Gilloire and S. Theodoridis, Complexity reduction in fast RLS transversal adaptive filters with application to acoustic echo cancellation, ICASSP92, pp. Ⅳ-37-Ⅳ-40
[30] D. T. M. Slock, On the convergence behavior of the LMS and the normalized LMS algorithm, IEEE Transactions on Signal Processing, September 1993, vol. 41, no. 9, pp. 2811-2825
[31] M. Tarrab and A. Feuer, Convergence and performance analysis of the normalized LMS algorithm with uncorrelated Gaussian data, IEEE Transactions on Information Theory, July 1988, vol. 34, no. 4, pp. 680-691
[32] J. Prado and E. Moulines, Frequency-domain adaptive filter with application to acoustic echo cancellation, Annals de Telecommunications, 1994, vol. 49, no. 7-8, pp. 414-428
[33] E. Eleftheriou and D. D. Falconer, Tracking properties and steady-state performance of
RLS adaptive filter algorithms,IEEE Transactions on Acoustics,Speech,and Signal Processing,October 1986,vol.ASSP-34,no.5,pp.1097-1110
[34] D.R.Morgan and J.C.Thi,A delayless subband adaptive filter architecture,IEEE Transactions on Signal Processing,August 1995,vol.43,no.8,pp.1819-1830
[35] A.Gilloire,Experiments with subband acoustic echo cancelers for teleconferencing, Proceedings of ICASSP,Dallas,April 1987,pp.2141-2144
[36] K.Ashihara,K.Nishikawa and H.Kiya,Improvement of convergence speed for subband adaptive digital filters using multirate repeating method,Proceedings of ICASSP,Detroit, May 1995,pp.989-992
[37] R.Martin and J.Altenhoner,Coupled adaptive filters for acoustic echo control and noise reduction,Proceedings of ICASSP,Detroit,May 1995,pp.3043-3046
[38] V.A.Margo,D.M.Etter and J.H.Gross,Multiple short length adaptive filters for time varying echo cancellation,ICASSP 93,pp.1-161-164
[39] R.Storn,Echo cancellation techniques for multimedia applications-a Survey,TR-96-046, 1996
[40] J.D.Johnston,A filter family designed for use in quadrature mirror filter banks, ICASSP80,pp.291-294
[41] H.Nussbaumer and M.Veterli,Computaionally efficient QMF filter banks,ICASSP84,pp. 11． 3． 1-11． 3． 4
[42] W.Kellermann,Analysis and design of multirate systems for cancellation of acoustical echos,ICASSP88,pp.2570-2573
[43] F.A.Reed,P.L.Feintuch,A comparison of LMS adaptive cancellers implemented in the frequency domain and the time domain,IEEE Trans,on Circuits and Systems,June 1981, vol.CAS-28,no.6
[44] M.R.Asharif,T.Hayashi and K.Yamashita,Correlation LMS algorithm and its application to double-talk echo canceling,Electronics Letters,Feb.1999,vol.18,no.5,pp. 1062-1068
[45] K.Fujii and J.Ohga,A fast adaptive algorithm suitable for acoustic echo canceller,IEICE Trans.Fundamentals,Nov.1992,vol.E75-A
[46] T.Ganseler,A double talk resistant subband echo canceller,Signal Processing,1998, vol.65,pp.89-101
[47] T. Ganseler, M. Hansson, C. Ivarsson and G. Salomonsson, A double-talk detector based on coherence, IEEE Trans. on Communications, Nov. 1996, vol. 44 no. 11
[48] 藤井健作、大贺寿郎，音响有用推定误差安定的确保方法，电子情报通信学会论文志，EA-97，Nov．1997，pp．53-60
[49] H. Ye and B. Wu, A new double-talk detection algorithm based on the orthogonality theorem, IEEE Trans. on Communications, Nov. 1991, vol. 39, no. 11
[50] N. Nilsson, An echo canceller with frequency dependent NLP attenuation, Ericsson Radio Systems AB, Kista Sweden, Jun. 1998
[51] J.R. Deller, J. G. Proakis and J. H. L. Hansen, Discrete-time processing of speech signals, Macmillan Publishing Company, New York, 1993
[52] 宗孔德，多抽样率信号处理，清华大学出版社，1996
[53] 秦前清等，实用小波分析，西安电子科技大学出版社，1993
[54] B. Ayad, G. Faucon and R. L. Bouquin-Jeannes, Optimization of a noise reduction preprocessing in an acoustic echo and noise controller, Proceedings of ICASSP, Atlanta, May 1996, pp. 953-956