本論文提出應用麥克風陣列技術於聲學迴聲消除系統的方法。不同於單一聲道的迴聲消除系統一般都使用複雜的適應性演算法，本論文利用麥克風陣列訊號處理技術，設計麥克風陣列的指向性，加強從近端通話者方向過來的語音，並且同時抑制從空間響應傳遞過來的迴聲，以達到在空間上濾波的效果。
然而存在著三個系統誤差會嚴重地影響麥克風陣列技術的穩健性與指向性，此三個系統誤差分別為(i)麥克風陣列單體之間的頻率響應不一致、(ii)麥克風陣列其真實的聲學中心與預計擺放位置的誤差以及(iii)語音訊號傳遞到麥克風陣列不在陣列指向性主軸內的誤差，而且當麥克風尺寸縮小時，這些影響會更劇烈。另外，麥克風陣列技術也可利用廣義旁葉瓣消除器，此演算法架構有效地消除主軸以外的干擾源並且自動地補償與校正系統誤差，模擬與語者辨識實驗皆可驗證陣列消除噪音的能力。
為了更符合現實上的應用，本論文更利用分頻帶訊號處理，在不同頻帶給予其適當的參數值，不但可以增加迴聲消除的能力與減少計算複雜度。實驗數據顯示出陣列訊號處理確實可以提升迴聲消除與語音加強的能力。

Acoustic echo that can substantially undermine speech quality is one of the key issues one must address in practical telecommunication systems. Distinct from conventional mono-channel acoustic echo cancellation (AEC) techniques, this study proposes an acoustic echo jammer (AEJ) that takes advantage of beamforming to nullify the echo path. Highly directional microphone array is designed to focus on the near-end speaker and suppress the echo from the far-end as well as noise and interference from the back ground. However, various forms of system deviations, channel mismatch, sensor location error, and pointing error, severely degrade the performance of array beamformer. To cope with these deviations, generalized sidelobe canceller (GSC) not only adaptively compensates and calibrates the system errors but also suppresses the interference of sidelobe. Simulations and Automatic Speech Recognition (ASR) experiment imply the speech quality enhancement of array beamformer. Objective tests of echo cancellation performance metric ERLE (Echo Return Loss Enhancement) and Perceptual Evaluation of Speech Quality (PESQ) are calculated to compare several implementation approaches. In addition, subband (SB) implementation is employed to improve the processing efficiency. Experimental investigations show that with the beamformer, the cancellation performance can be significantly increased, as compared to conventional AEC.

摘 要 I
ABSTRACT II
誌 謝 III
LIST OF TABLES VI
LIST OF FIGURES VII
I. INTRODUCTION 1
II. ADAPTIVE FILTERING 3
A. Adaptive Systems 3
B. Adaptive Algorithms 4
1. Least Mean Square Algorithm 5
2. Normalized Least Mean Square Algorithm 6
3. Proportionate Normalized Least Mean Square Algorithm 6
4. Affine Projection Algorithm 7
C. Sub-band Filtering 8
D. Performance Measure of Acoustic Echo Cancellation 9
III. NONLINEAR ACOUSTIC ECHO CANCELLATION 10
A. Second-Order Volterra Filter 10
B. Hammerstein Model 11
C. Modified Nonlinear Echo Cancellation System 13
D. Simulations and Experiments 14
IV. FARFIELD ARRAY SIGNAL PROCESSING 15
A. Farfield Array Model and Performance Measures 15
B. Beamformer Design 21
1. Super Directive Beamformer 21
2. Convex Optimization 24
C. Generalized Sidelobe Canceller 26
1. Griffiths-Jim Beamformer 27
2. Singular Value Decomposition 27
3. Leaky Adaptive Filter Structure 28
D. Simulations and Experiments 29
1. Simulations of Interference Suppression 29
2. Microphone Mismatch Test 30
3. Automatic Speech Recognition 32
V. ARRAY-BASED ADAPTIVE ACOUSTIC ECHO JAMMER 33
A. Array Echo Jammer 33
B. Sub-band GSC 34
C. Simulations and Experiments 34
VI. CONCLUSIONS 37
REFERENCES 39

1 S. M. Kuo and D. R. Morgan, “Active Noise Control System,” John-Wiley, New York (1996).
2 K. A. Lee, W. S. Gan, and S. M. Kuo, “Subband Adaptive Filtering: Theory and Implementation,” Wiley, (2009).
3 S. L. Gay and J. Benesty, “Acoustic Signal Processing for Telecommunication,” Kluwer Academic, London (2000).
4 E. Hänsler and G. Schmidt, Acoustic Echo and Noise Control: A Practical Approach, Wiley, New York (2004).
5 J. B. Seo, K. J. Kim, and S. W. Nam, “Nonlinear Acoustic Echo Cancellation Using Volterra Filtering with a Variable Step-Size GS-PAP Algorithm,” World Academy of Science, Engineering and Technology, Vol. 57, pp. 59-62 (2009).
6 J. Benesty and S. L. Gay, “An improved PNLMS algorithm,” IEEE International Conference on Acoust. Speech, and Signal Processing, Orlando, May (2002).
7 K. Fai, C. Yiu, Y. Lu, C. H. Ho, W. Luk, J. Huoc, and S. Nordholm, “Reconfigurable FPGA-based robust frequency-domain echo canceller with applications to voice control device,” Digital Signal Processing, Vol. 22, pp. 376-390 (2011).
8 J. Q. Huo, K. F. C. Yiu, S. Nordholm, and K. L. Teo, “A robust transform domain echo canceller employing a parallel filter structure,” Signal Processing, Vol. 86, No. 12, pp. 3752–3760 (2006).
9 Y. P. Lin and P. P. Vaidyanatjan, “A kaiser window approach for the design of prototype filter of cosine modulated filterbanks,” IEEE, Signal Processing Letters, Vol. 5, pp. 132-134 (1998).
10 D. Zhou and V. De Brunner, “A New Active Noise Control Algorithm That Requires No Secondary Path Identification Based on the SPR Property,” IEEE Transactions on Signal Processing, Vol. 55, pp.1719-1729 (2007).
11 C. Breining, P. Dreiscitel, E. Hansler, A. Mader, B. Nitsch, H. Puder, T. Schertler, G. Schmidt, and J. Tilp, “Acoustic echo control. An application of very-high-order adaptive filters,” IEEE Signal Processing Magazine, Vol.16, pp. 42-69 (1999)
12 A. Stenger, L. Trautmann and R. Rabenstein, “Nonlinear acoustic echo cancellation with 2nd order adaptive Volterra filters,” IEEE International Conference on Acoustics, Speech, Signal Processing, Phoenix, Vol. 2, pp. 877-880 (1999).
13 A. Stenger and W. Kellermann, “Adaptation of a memoryless preprocessor for nonlinear acoustic echo cancelling,” Signal Processing, Vol. 80, No. 9, pp. 1747-1760 (2000).
14 E. J. Thomas, “Some considerations on the application of the Volterra representation of nonlinear networks to adaptive echo canceller,” Bell System Technical Journal, Vol. 50, No. 8, pp. 2797-2805 (1971).
15 P. P. Vaidyanathan, “Multirate Systems and Filter Banks,” Prentice-Hall PTR, Englewood Cliffs, NJ (1992).
16 ITU-T Recommendation G.168, “Digital network echo cancellers,” International Telecommunication Union, Geneva, Switzerland, 132pages (2012).
17 A. Stenger, R. Rabenstein, “An Acoustic Echo Canceller With Compensation of Nonlinearities,” Proc. EUSIPCO 98, Isle of Rhodes, Greece, pp. 969-972 (1998).
18 F. Kuech and W. Kellermann, “Proportionate NLMS algorithm for second-order Volterra filters and its application to nonlinear echo cancellation,” in Proc. Workshop on Acoustic Echo and Noise Control, Kyoto, pp. 75-78 (2003).
19 ITU-T P.56, “Objective measurement of active speech level,” ITU-T Recommendation ( 1993).
20 W. Klippel, “Distortion Analyzer – a New Tool for Assessing and Improving Electrodynamic Transducer,” presented at the 108th Convention of the Audio Eng. Soc., Paris, February 19-22, (2000).
21 M. R. Bai, J. G. Ih, and J. Benesty, “Acoustic Array Systems: Thoery, Implementation, and Application,” John-Wiley IEEE, New York (2013).
22 J. Mattingley and S. Boyd, “Real-time convex optimization in signal processing,” IEEE Signal Process. Mag. 27, 50-61 (2010).
23 S. Boyd and L. Vandenberghe, Convex optimization, (Cambridge University Press, New York, 2004), Chap. 1-7..
24 M. Grant and S. Boyd, cvx, Version 1.21 MATLAB software for disciplined convex programming available at http://cvxr.com/cvx (last viewed February 14, 2012).
25 M. Brandstein and D. Ward, “Microphone Arrays: Signal Techniques and Applications,” Springer, Berlin (2001).
26 L. J. Griffiths and C.W. Jim, “An alternative approach to linear constrained adaptive beamforming,” IEEE Trans. Antennas Propagat., pp.27-34, Jan. 1982.
27 W. Herbordt, H. Burchner, and W. Kellerman, “An Acoustic Human-Machine Front-End for Multimedia Applications,” EURASIP Journal on Applied Signal Processing, 1, 21-31(2003).
28 T. Yardibia, C. Bahrb, N. Zawodnyb, F. Liub, L.N. Cattafesta and J. Li, “Uncertainty analysis of the standard delay-and-sum beamformer and array calibration,” J. Sound Vibrat. 329(13), 2654-2682 (2010).
29 P. Castellini and M. Martarelli, “Acoustic beamforming: Analysis of uncertainty and metrological performances,” Mech. Syst. Signal Pr. 22(3), 672-692 (2008).
30 J. G. Wilpon, L. R. Rabiner, C. H. Lee and E. R. Goldman, “Automatic recognition of keyword in unconstrained speech using hidden Markov models,” IEEE Trans. Acoustics, Speech, Signal Proc., 38(11), 1870-1878 ITU-T Recomm (1990).
31 ITU-T Recommendation P.862, “Perceptual evaluation of speech quality (PESQ): An objective method for end-to-end speech quality assessment of narrow-band telephone networks and speech codecs,” International Telecommunication Union, Geneva, Switzerland, 21pages (2001).
32 ITU-T Recommendation P.862.2, “Wideband extension to Recommendation P.862 for the assessment of wideband telephone networks and speech codecs,” International Telecommunication Union, Geneva, Switzerland, 4 pages (2007).