The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지
DOI QR코드

DOI QR Code

An Adaptive Time Delay Estimation Method Based on Canonical Correlation Analysis

정준형 상관 분석을 이용한 적응 시간 지연 추정에 관한 연구

  • Lim, Jun-Seok (Department of Defense Systems Engineering, Sejong University) ;
  • Hong, Wooyoung (Department of Electronics Engineering, Sejong University)
  • Received : 2013.07.22
  • Accepted : 2013.08.27
  • Published : 2013.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

The localization of sources has a numerous number of applications. To estimate the position of sources, the relative delay between two or more received signals for the direct signal must be determined. Although the generalized cross-correlation method is the most popular technique, an approach based on eigenvalue decomposition (EVD) is also popular one, which utilizes an eigenvector of the minimum eigenvalue. The performance of the eigenvalue decomposition (EVD) based method degrades in the low SNR and the correlated environments, because it is difficult to select a single eigenvector for the minimum eigenvalue. In this paper, we propose a new adaptive algorithm based on Canonical Correlation Analysis (CCA) in order to extend the operation range to the lower SNR and the correlation environments. The proposed algorithm uses the eigenvector corresponding to the maximum eigenvalue in the generalized eigenvalue decomposition (GEVD). The estimated eigenvector contains all the information that we need for time delay estimation. We have performed simulations with uncorrelated and correlated noise for several SNRs, showing that the CCA based algorithm can estimate the time delays more accurately than the adaptive EVD algorithm.

음원 위치 추정은 여러 방면에서 쓰임이 있는 응용 기술이다. 음원의 위치를 추정하기 위한 기본 기법 중에는 시간 지연 추정 기법이 있다. 이 기법에선 음원의 위치를 추정하기 위해서 두 개 또는 그 이상의 수신기에 들어오는 신호간의 상대적 시간 지연을 알아내야 한다. 시간 지연 추정 기법에는 GCC (Generalized Cross-Correlation) 대표적이지만, 최소 고유치에 대응하는 고유 벡터를 이용하는 방법도 많이 쓰인다. 이 방법은 최소 고유치에 해당하는 고유벡터를 이용한다. 최소 고유치에 대응하는 고유 벡터를 이용하는 방법은 낮은 신호 대 잡음비 환경에서나 상관도가 있는 잡음환경에서, 최소 고유치에 해당하는 고유 벡터를 추정하는데 어려움이 있어서, 성능이 떨어진다. 본 논문에서는 정준형 상관 분석 (CCA)를 이용한 새 기법을 제안한다. 이 방법은 일반 고유치 분해 중에서 최대 고유치에 대응하는 고유벡터를 사용한다. 따라서 추정에 사용하는 고유벡터는 시간 지연 추정에 필요한 정보가 충분히 들어있다. 본 논문에서는 여러 서로 다른 신호 대 잡음비 환경 하에서 상관도가 없는 경우와 상관도가 있는 경우의 잡음 에 대해 비교 모의실험을 하였고, 이 비교 실험을 통하여 얻는 데이터를 통해서 제안한 CCA 기반 알고리즘이 기존 최소 고유치에 해당하는 고유벡터를 사용하는 시간 지연 추정법의 성능보다 더 우수하다는 것을 보인다.

Keywords

References

  1. Havelock D, Kuwano S, Vorlander M., Handbook of signal processing in acoustics (Springer, Berlin, 2008), pp. 33-52.
  2. Ferreira JF, Pinho C, Dias J., "Implementation and calibration of a Bayesian binaural system for 3D localisation," IEEE international conference on robotics and biomimetics 2008, 1722-727 (2008).
  3. Tiana-Roig E, Jacobsen F, Fernandez Grande E., "Beamforming with a circular microphone array for localization of environmental noise sources," J. Acoust. Soc. Am.128, 3535-42 (2010). https://doi.org/10.1121/1.3500669
  4. May T, van de Par S, Kohlrausch A., "A probabilistic model for robust localization based on a binaural auditory front-end," IEEE Trans Audio Speech Lang Process. 19, 1-13 (2011). https://doi.org/10.1109/TASL.2010.2042128
  5. G.C.Carter, Coherence and Time Delay Estimation: An Applied Tutorial for Research, Development, Test and Evaluation Engineers (IEEE press, NewYork, 1993), pp. 1-28.
  6. C. H. Knapp and G. C. Carter, "The generalized correlation method for estimation of time delay," IEEE Trans. Acoust., Speech, Signal Process. 24, 320-327, (1976). https://doi.org/10.1109/TASSP.1976.1162830
  7. B. Champagne, S. Bedard, and A. Stephenne, "Performance of timedelay estimation in presence of room reverberation," IEEE Trans. Speech Audio Processing, 4, 148-152 (1996). https://doi.org/10.1109/89.486067
  8. Jacob Benesty, "Adaptive eigenvalue decomposition algorithm for passive acoustic source localization," J. Acoust. Soc. Am. 107, 384-391 (2000). https://doi.org/10.1121/1.428310
  9. S. Doclo, M. Moonen, "Robust adaptive time delay estimation for speaker localization in noisy and reverberant acoustic environments," EURASIP J. Appl. Signal Process., 11, 1110-1124 (2003).
  10. H. Hotelling, "Relations between two sets of variates". Biometrika, 28, 321-377 (1936). https://doi.org/10.1093/biomet/28.3-4.321
  11. Dogandzic, A., & Nehorai, A., "Finite-length MIMO equalization using canonical correlation analysis," IEEE Transactions on Signal Processing, 50, 984-989 (2002). https://doi.org/10.1109/78.992151
  12. Dogandzic, A., & Nehorai, A., "Generalized multivariate analysis of variance; A unified framework for signal processing in correlated noise," IEEE Signal Processing Magazine, 20, 39-54 (2003).
  13. J. Via, I. Santamaria, and J. Perez, "A learning algorithm for adaptive canonical correlation analysis of several data set," Neural Network, 20, 139-152 (2007). https://doi.org/10.1016/j.neunet.2006.09.011
  14. Friman, O., Borga, M., Lundberg, P., and Knutsson, H., "Adaptive analysis of fMRI data," Neuroimage, 19, 837-845 (2003). https://doi.org/10.1016/S1053-8119(03)00077-6
  15. Javier Via, Ignacio Santamaria and Jesus Perez, "A learning algorithm for adaptive canonical correlation analysis of several data sets," Neural Networks, 20, 139-152 (2007). https://doi.org/10.1016/j.neunet.2006.09.011
  16. M. Borga, Learning Multidimensional Signal Processing, (Ph.D. thesis, Linkoping University, Linkoping, Sweden, 1998).
  17. Z. Chen, S. Haykin, J.J. Eggermont and S. Becker. Correlative Learning, A Basis for Brain and Adaptive Systems (John Wiley & Sons, Hoboken, N.J., 2007), pp. 130-157.
  18. M. Moonen, P. Van Dooren, and J. Vandewalle, "A systolic algorithm for QSVD updating," Signal Processing, 25, 203-213 (1991). https://doi.org/10.1016/0165-1684(91)90063-O
  19. S. Haykin, Adaptive Filter Theory , Fourth edition (Prentice Hall, New Jersey, 2002), pp. 231-319.
  20. H. C. So, "On time delay estimation using an FIR filter," Signal Processing, 81, 1777-1782 ( 2001). https://doi.org/10.1016/S0165-1684(01)00098-6

Cited by

  1. Time Delay Estimation Using LASSO (Least Absolute Selection and Shrinkage Operator) vol.39B, pp.10, 2014, https://doi.org/10.7840/kics.2014.39B.10.715