Recovery of LSP Coefficient in VoIP Systems using Evolving Takagi-Sugeno Fuzzy Models

In order to deliver real time, high quality voice services in packet based voice system (e.g. voice over Internet protocol, VoIP) system designers must tackle inherent quality problems related to possible packet loss. To combat the inevitable speech quality deterioration resulting from the loss of transmitted packets of speech information, techniques that provide estimates of the lost information that is needed by the speech recovery process are of considerable interest. Furthermore, in VoIP systems employing linear predictive coding (LPC) based speech coders, a significant percentage of the coded speech information represent the values of LPC coefficients and thus a new approach for estimating missing LPC filter coefficients is presented in this paper. This approach employs a new formulation of LSP recovery system architecture where evolving fuzzy rule-based models and particularly so-called evolving Takagi-Sugeno models are deployed to generate the required estimates of missing LSPs. The proposed missing parameters estimation technique is generic and initial experimental results demonstrate its considerable potential in improving the quality of LPC based decoded speech in VoIP applications

[1]  Geoff Huston,et al.  Quality of Service: Delivering QoS on the Internet and in Corporate Networks , 1998 .

[2]  Plamen Angelov,et al.  Evolving Rule-Based Models: A Tool For Design Of Flexible Adaptive Systems , 2002 .

[3]  Stephen L. Chiu,et al.  Fuzzy Model Identification Based on Cluster Estimation , 1994, J. Intell. Fuzzy Syst..

[4]  C. Xydeas,et al.  Recovery of LSP Coefficients in VoIP Systems , 2005, 2005 5th International Conference on Information Communications & Signal Processing.

[5]  Thomas P. Barnwell,et al.  Speech Coding: A Computer Laboratory Textbook , 1995 .

[6]  Dimitar Filev,et al.  On-Line Evolution of Takagi-Sugeno Fuzzy Models , 2004, IFAC Proceedings Volumes.

[7]  Plamen P. Angelov,et al.  On-line identification of MIMO evolving Takagi- Sugeno fuzzy models , 2004, 2004 IEEE International Conference on Fuzzy Systems (IEEE Cat. No.04CH37542).

[8]  Chia-Feng Juang,et al.  A recurrent self-organizing neural fuzzy inference network , 1997, Proceedings of 6th International Fuzzy Systems Conference.

[9]  Ronald R. Yager,et al.  Learning of Fuzzy Rules by Mountain Clustering , 1992 .

[10]  D.P. Filev,et al.  An approach to online identification of Takagi-Sugeno fuzzy models , 2004, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).