A_PSQA: PESQ-like non-intrusive tool for QoE prediction in VoIP services
Authors:
- 1 June 2012
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15503607,p. 2124-2128
Abstract
Perceived speech quality, or Quality of Experience (QoE), is the key criteria for evaluating VoIP service. Most of the existing solutions are intrusive, in a sense where they require both the original and the transmitted audio sequences. These solutions give good estimation of the QoE, but they cannot be used in real-time. In fact, Service Provider and Network Provider are highly interested on automatic QoE estimation tool (without intrusion) in order to monitor and control the perceived quality of their VoIP service. In this paper, we present a new perceived speech quality estimation tool, named ALICANTE 1 Pseudo Subjective Quality Assessment (A_PSQA) for two widely VoIP codecs, iLBC and Speex. A_PSQA is a non-intrusive method, which relies on Random Neural Network (RNN) approach to learn the nonlinear relation between network parameters and the perceived user QoE. Furthermore, to avoid costly and time-consuming subjective tests, we used a well-known intrusive method ITU-T's Perceptual Evaluation Quality (PESQ) to estimate the MOS. Obtained results show that A_PSQA is able to estimate the MOS like PESQ, while being non-intrusive. Besides, A_PSQA's results are compared with two non-intrusive (IQX and E-Model) methods, where A_PSQA shows the highest correlation with PESQ estimation than all others methods.Keywords
This publication has 13 references indexed in Scilit:
- Performance evaluation of real-time speech through a packet network: a random neural networks-based approachPerformance Evaluation, 2004
- Prediction of perceived conversational speech quality and effects of playout buffer algorithms2004
- Perceived speech quality prediction for voice over IP-based networks2003
- A passive method for monitoring voice-over-IP call quality with ITU-T objective speech quality measurement methods2003
- Testing the ability of speech recognizers to measure the effectiveness of encoding algorithms for digital speech transmission2003
- Neural network-based voice quality measurement technique2003
- Integrating networks measurements and speech quality subjective scores for control purposes2002
- Voice over IP performance monitoringACM SIGCOMM Computer Communication Review, 2001
- Measurement and modelling of the temporal dependence in packet loss1999
- Learning in the Recurrent Random Neural NetworkNeural Computation, 1993