Title :
Replacement of missing speech parameters using Markov chain estimates
Author_Institution :
U.S. Dept. of Defense, Fort Meade, MD, USA
Abstract :
Packet-switched networks are popular for voice communication although many packet-switching protocols cannot guarantee timely delivery of every packet. This paper describes a technique for replacing lost/missing voice parameters at the receiver using the mixture transition distribution to estimate Markov chain probabilities of states consisting of quantized voice coding parameters
Keywords :
Markov processes; compensation; digital communication; packet switching; prediction theory; probability; speech coding; voice communication; Markov chain estimates; Markov chain prediction; Markov chain probabilities; lossy transmission system; lost voice parameters; missing speech parameters replacement; missing voice frames; missing voice parameters; mixture transition distribution; packet-switched networks; quantized voice coding parameters; voice communication; Access protocols; Computer errors; Computer networks; Delay effects; IP networks; Internet telephony; Probability; Speech; State estimation; Wide area networks;
Conference_Titel :
Circuits and Systems, 2000. Proceedings of the 43rd IEEE Midwest Symposium on
Conference_Location :
Lansing, MI
Print_ISBN :
0-7803-6475-9
DOI :
10.1109/MWSCAS.2000.952822
