Prediction of Speech Quality Based on Resilient Backpropagation Artificial Neural Network

• Published in: Wireless personal communications Vol. 96; no. 4; pp. 5375 - 5389
• Main Authors: Orcik, Lukas, Voznak, Miroslav, Rozhon, Jan, Rezac, Filip, Slachta, Jiri, Toral-Cruz, Homero, Lin, Jerry Chun-Wei
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.10.2017; Springer Nature B.V
• Subjects: Artificial neural networks; Back propagation; Back propagation networks; Communications Engineering; Computer Communication Networks; Data analysis; Engineering; Environmental monitoring; Internet telephony; Markov chains; Networks; Neural networks; Quality; Quality assessment; Signal,Image and Speech Processing; Speech; Telephony; Markov models; Neural networks; Speech quality; Network probes
• ISSN: 0929-6212; 1572-834X
• DOI: 10.1007/s11277-016-3746-2

The paper presents a system for monitoring and assessment the speech quality in the IP telephony infrastructures using modular probes. The probes are placed at key nodes in the network where aggregating packet loss data. The system dynamically measures speech quality and results are collected on a central server. For data analysis we applied four-state Markov model for modeling the impact of network impairments on speech quality, afterwards, the resilient back propagation (Rprop) algorithm was used to train a neural network. Information about the speech quality are displayed in the form of automatically generated graphs and tables. The proposed solution has been tested with selected codecs and further generalizes the already presented concepts of the speech quality estimation in the IP environment.