Abstract
Acoustic echo cancellation in hands-free telephones can be regarded as adaptive system identification with several hundreds of filter coefficients in very noisy environment and with unstationary non-white excitation. Its performance is highly dependent on the choice of the time variant adaptation step size. The step size can be controlled with the help of various methods which estimate statistical characteristics of the measurable signals, but which are not very reliable and should be combined to improve the resulting step size. In this paper, we present some results of such a combination by means of a multilayer perceptron, and show how different cost functions and training situations affect the performance of the echo canceller.
Preview
Unable to display preview. Download preview PDF.
References
P. Heitkämper: Optimization of an acoustic echo canceller combined with adaptive gain control, Proc. ICASSP-95, Detroit, Michigan, 1995, pp. 3047–3050
J. Marx: Akustische Aspekte der Echokompensation in Freisprecheinrichtungen, VDI-Fortschritt-Berichte, Reihe 10, Nr. 400, Düsseldorf, 1996
R. Frenzel: Freisprechen in gestörter Umgebung, VDI-Fortschritt-Berichte, Reihe 10, Nr. 228, Düsseldorf, Germany, 1992
S. Yamamoto, S. Kitayama: An Adaptive Echo Canceller with Variable Step Gain Method, Trans. IECE of Japan vol. E 65, no. 1, 1982, pp. 1–8.
International Telecommunication Union: General characteristics of international telephone connections and international telephone circuits — acoustic echo controllers. ITU-T Recommendation G.167, 1993.
C. Breining: On Applying a Neural Network for Step size Control in Echo Cancellation, submitted to Int. Workshop on Acoustic Echo and Noise Control 97, London, September 1997
D. E. Rumelhart, J. L. McClelland: Parallel Distributed Processing,vol. 1, MIT Press, Cambridge, Massachusetts, 1986
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Breining, C. (1997). On using MLPs for step size control in echo cancellation for hands-free telephone sets. In: Gerstner, W., Germond, A., Hasler, M., Nicoud, JD. (eds) Artificial Neural Networks — ICANN'97. ICANN 1997. Lecture Notes in Computer Science, vol 1327. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0020260
Download citation
DOI: https://doi.org/10.1007/BFb0020260
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63631-1
Online ISBN: 978-3-540-69620-9
eBook Packages: Springer Book Archive