Nonlinear Acoustic Echo-Free System

Authors

  • Deshpande Tanavi A. and Dube R. R. Author

Keywords:

Acoustic echo canceller, adaptive algorithm, NLMS, non linear transformation, raised cosine function.

Abstract

Acoustic Echo Cancellation (AEC) is one of the  
most popular applications for adaptive filters. The      
goal of the adaptive algorithms is to identify the acoustic echo path between a loudspeaker and a microphone placed in an acoustic enclosure (i.e., the room acoustic impulse response) and generate the estimate of the acoustic echo that is removed from 
the output of the microphone. Most of the adaptive echo cancellation techniques proposed in literature are based on an assumption that the echo path is linear. However, recent research has shown that the acoustic echo path is better to be modeled as a nonlinear system due to the nonlinear characteristic of a loudspeaker[1].The nonlinearity of a power amplifier or loudspeaker in a large-signal situation gives rise to 
a nonlinear distortion of acoustic signal. A conventional 
acoustic echo canceller using linear adaptive filters is not able to eliminate the nonlinear echo component. Here, a novel nonlinear echo cancellation technique is presented by using a nonlinear transformation in conjunction with a conventional linear adaptive filter. The nonlinear transformation is derived from a raised-cosine function and is exploited to compensate for 
the nonlinearity of a loudspeaker. The algorithms which were derivates out of the Wiener Filter (like LMS, NLMS, steepest descent, RLS) in contrast to the Genetic algorithms (GAs) is of special interest[6]. With the comparison, NLMS algorithm is better solution  considering trade-offs. The transformation 
parameters are updated using the normalized least mean square algorithm according to the unknown nonlinear characteristic of the loudspeaker.

Downloads

Download data is not yet available.

Downloads

Published

06.07.2012

How to Cite

Nonlinear Acoustic Echo-Free System. (2012). International Journal of Information and Electronics Engineering, 2(4), 635-639. https://ijiee.org/index.php/ijiee/article/view/191