VoIP quality measurement: Enhanced E-model using bias factor

Abstract

To enhance the standard E-model, this paper proposes the new factor called Bias factor. Based-on the Thai environment, this factor can be computed by subtracting the subjective test results using conversation tests with native Thai users from the objective test results using an E-model tool. Both E-mode tests and conversation tests were conducted with the same VoIP system and test conditions. Then, the enhanced E-model using the Thai bias factor was evaluated with the test set from other groups of native Thai users. After evaluation, it has been found that the enhanced E-model can gain higher confidence. It delivers improved accuracy and reliability at approximately more than 20% when compared to the standard E-model.

VoIP quality measurement: subjective VoIP quality estimation model for G.711 and G.729 based on native Thai users

“…Also, in another work, Ren et al proposed the language impairment factor I l for Chinese referring to packet loss effects [57]. Recently, Assem et al [40] proposed the improvement of the simplified E-model for four codecs, G.711, G.723.1 (5.3 kbps), G.726 (4 kbps), and G.729A based on MOS-LQO from PESQ, whereas Daengsi and Wuttidittachotti [58] proposed the enhanced E-model using bias factor, which is the representative for cultural factor and language factor. Moreover, there are previous works that proposed a new VoIP quality evaluation/prediction models.…”

QoE Modeling for Voice over IP: Simplified E-model Enhancement Utilizing the Subjective MOS Prediction Model: A Case of G.729 and Thai Users

“…Daengsi & Wuttidittachotti (2013) [43] The Thai bias factor and MOSCQE-TH was proposed for G.711 as shown in (18) and (19), where Thai bias factor is the factor obtained from subjective tests with a groups of native Thais, MOSCQE stands for MOS from an E-model tool, K is 5 for Thai, D stands for delay (s); 0s<delay<0.8s only, L stands for loss rate; 0 %<loss<10 % only, MOSCQE-TH stands for objective-MOS from E-model with Thai bias factor (BiasTH). BiasTH = -K(56.12-255.4L+88.48D-901DL-179.88D 2 )/10 The improved the simplified E-model was proposed for several codecs (e.g.…”