Journal of Innovation in Electronics and Communication Engineering
  • Year: 2016
  • Volume: 6
  • Issue: 2

Impact of Enhanced Full Rate Speech Codec on the Performance of ASR over GSM Networks

  • Author:
  • Mohan Dholvan1, Anitha Sheela Kancharla2, Jagadheesh Samala3
  • Total Page Count: 4
  • Page Number: 21 to 24

1Associate Professor, Department of Electronics and Computer Engineering, Sreenidhi Institute of Science and Technology, Hyderabad. Email: mohan.aryan19@sreenidhi.edu.in

2Associate Professor, Department of Electronics Communications Engg., JNTUCE, Hyderabad, Telangana, India. kanithasheela@gmail.com

3Assistant Professor, Department of Electronics and Computer Engineering, Sreenidhi Institute of Science and Technology, Hyderabad. jagadeeshsamala@gmail.com

Online published on 27 June, 2017.

Abstract

This paper explains how narrow band Enhanced Full Rate speech and channel coding algorithms are applied to evaluate the efficiency and mean opinion score (MoS) of Automatic Speech Recognition (ASR) over GSM Network and it also illustrates how to estimate the ASR accuracy using CMU SPHINX toolkit and TIMIT speech database (8 KHz raw files). MoS values are computed by using PESQ P.862. In this work we have used Gaussian Minimum Shift keying (GMSK) modulation for the speech transmission and channel noise is added by considering Raleigh fading channel models. Later, we showcase and analyze the results obtained in this work

Keywords

GSM-EFR, CMU Sphinx, TIMIT DATABASE, GMSK, ASR, MoS