SELECTING TEST SIGNALS FOR SUCCESSFUL IMPAIRMENT CLASSIFICATION IN VOIP SYSTEMS

Doris Bao, Luca De Vito, Sergio Rapuano
Abstract:
The paper proposes a metrological analysis for selecting the test signals to be used with a method for the automatic classification of impairments in VoIP telecommunication systems. The method is able to identify the VoIP network impairments which decrease voice quality, by analyzing the power of a test signal during the call. Test signals, considered in this work, comprise speech-like signals, non- speech-like deterministic signals and non- speech-like stochastic signals, taken either from standard specifications or from literature papers on voice quality assessment. In order to evaluate the optimal test signals, two network emulators have been used to add the impairments in a controlled test network. The selected test signals improve significantly the method performance in terms of classification success rate and test duration.
Download:
IMEKO-TC4-2008-203.pdf
DOI:
-
Event details
IMEKO TC:
TC4
Event name:
Exploring New Frontiers of Instrumentation and Methods for Electrical and Electronic Measurements
Title:
XVIth IMEKO TC4 International Symposium on Electrical Measurements and Instrumentation (together with 13th IMEKO TC4 Workshop on ADC Modelling and Testing)
Place:
Florence, ITALY
Time:
22 September 2008 - 24 September 2008