Perceptual Evaluation Of Playout Buffer Algorithm For Enhancing Perceived Quality Of Voice Transmission Over Ip Network

TL;DR

This paper evaluates and improves playout buffer algorithms for VoIP to enhance perceived speech quality over IP networks with variable impairments, proposing a new algorithm that adapts better to network conditions.

Contribution

The paper introduces a modified playout buffer algorithm that optimizes speech quality by adapting to network impairments more effectively than existing methods.

Findings

The new algorithm improves perceived speech quality in various network conditions.

It outperforms existing algorithms in handling high delay variations.

The proposed method is efficient for real-time buffer implementation.

Abstract

Voice over Internet Protocol (VoIP) is a technology that allows you to make voice calls using a broadband Internet connection instead of a regular (or analog) phone line. Voice over Internet Protocol (VoIP) has led human speech to a new level, where conversation across continents can be much cheaper & faster. However, as IP networks are not designed for real-time applications, the network impairments such as packet loss, jitter and delay have a severe impact on speech quality. The playout buffer at the receiver side is used to compensate jitter at a trade-off of delay and loss. We found the characteristics of delay and loss are dependent on IP network and sudden variable delay (spike) often performs both regular and irregular characteristics. Different playout buffer algorithms can have different impacts on the achievement speech quality. It is important to design a playout buffer algorithm which can help achieve an optimum speech quality. In this paper, we investigate to the understanding how network impairments and existing adaptive buffer algorithms affect the speech quality and further to design a modified buffer algorithm to obtain an optimized voice quality. We conduct experiments to existing algorithms and compared their performance under different network conditions with high and low network delay variations. Preliminary results show that the new algorithm can enhance the perceived speech quality in most network conditions and it is more efficient and suitable for real buffer mechanism.