Packet Loss Concealment with Natural Variations using HMM

Packet loss concealment (PLC) at a receiver has a substantial effect on the speech quality in voice over IP. Most conventional PLC systems have largely relied upon variations of signal repetition and overlap-add interpolation which can produce speech signals that do not follow the larger overall statistical trends. in this paper, we demonstrate how hidden Markov models can be utilized to effect PLC based on statistical signal processing. In particular, we show how HMM-based PLC yields conditional density functions that can be utilized by various statistical estimation methods that produce signal parameter estimates that produce more natural variation than conventional PLC methods, thereby providing much better speech quality