期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2009
卷号：9
期号：8
页码：312-316
出版社：International Journal of Computer Science and Network Security
摘要：The nonlinearity of a power amplifier or loudspeaker in a large- signal situation gives rise to a nonlinear distortion of acoustic signal. 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. This paper deals with the reconstruction of a time varying input in a nonlinear environment. The time varying signal used in this work is a (TV-AR) (Time Varying Auto-Regressive) model with stochastic parameters and power variations. The synthetic signal generated by the (TV-AR) model presented here can be used to simulate the speech signal temporal variations. In order to equalize a nonlinear channel excited by such signal, we use a modified version of NLMS (Normalized Least Mean Square) algorithm which depends of instantaneous power of the (TV-AR) signal. The simulation results shows that this modified version of NLMS offers a better solution for equalizing a nonlinear channel excited by a time varying input.
关键词：nonlinear system; time varying signal; equalization; adaptive algorithms
