Prediction techniques applied to Differential Pulse-Code Modulation systems for encoding speech signals
Year: 1982
Type: dissertation
Abstract: Differential pulse-code modulation (DPCM) is an efficient digitization\ntechnique for encoding speech signals. The two principal components of\na DPCM system are the quantizer and the predictor, either or both of\nwhich can be adaptive. This thesis describes the investigation of\nvarious differential pulse-code modulation systems. Initially, fixed,\ni.e. time-invariant, predictors using long-term signal statistics of the\nspeech signal are examined. The performance of such a predictor in a\nDPCM system having a fixed quantizer is studied. Then by replacing the\nfixed predictor with one whose coefficients are calculated at fixed time\nintervals, the performance of the encoder, in terms of signal to noise\nratio (SNR), is improved by 3-5 dB. A further improvement of 2–3 dB in\nSNR is obtained when an adaptive quantizer is used in the DPCM system.\nHowever, the block adaptive predictor requires the transmission of\nprediction coefficients to the receiving end at the expense of an increase\nin the channel band width. [Continues.]
Author: C. Evci
Language: English
Topic: Speech and Audio Processing
Subfield: Signal Processing
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2778284618