Adaptive Resource Utilization and Quality Management in VoIP Networks with Quantitative Relationship between QoE and QoS Using HMM

Abstract

Adaptive quality management schemes in Voice over IP (VoIP) Network have a constructive impact on the improvement of voice quality level received by end-users. These schemes are used for persistent monitoring of quality level leading to better management of network services. In this paper, a model is introduced to improve the voice quality level and optimize resource utilization. In the proposed model, the VoIP network makes the decision for configuration adaptively and modifies itself by means of Hidden Markov Model (HMM). Depending on the quality level and consumed bandwidth evaluated in the receiver side this adaptive self-modification takes place. During this procedure, acceptable weights for Quality and Quality of Service (QoS) parameters, namely, End-to-End Delay, Packet Loss, and Jitter are determined. These acquired weights help HMM model predict next optimum state properly. Another important objective of this work is to find a relation between Quality of Experience (in terms of Mean Opinion Score) and Quality of Service (in terms of End-to-End Delay and Jitter) parameters. This mapping provides a tangible understanding of how QoS parameters affect MOS. Simulation results demonstrate that the proposed technique lessens the impact of jitter and packet loss on voice quality level.