On predictive quantizing schemes

This paper analyzes the performance of various predictive quantizing schemes both for noiseless and noisy channels. The fidelity criterion used to define optimum performance is that of minimum mean-squared error. The first part of this paper compares differential pulse code modulation (DPCM) with a system that lacks the feedback around the quantizer. Such a system (that is called D∗PCM in this paper) is actually a pulse code modulation (PCM) system with a prefilter and a postfilter. In the second part of this paper a noise-feedback coding structure is used as a framework for a unified analysis of predictive quantizing schemes with a frequency-weighted mean-squared error as the performance criterion. The last part of this paper extends the analysis to include the effects of channel transmission errors on the overall performance of these predictive quantizing schemes. It is shown that DPCM and D∗PCM when appropriately optimized are less sensitive to channel errors than PCM, and that the performances of DPCM and D∗PCM are almost identical in the case of high bit-error rates.

[1]  A. Haddad,et al.  Some Properties of a Predictive Quantizing System , 1970 .

[2]  Jr. J. Spilker.  Theoretical Bounds on the Performance of Sampled Data Communications Systems , 1960 .

[3]  B. R. Narayana Murthy.  OPTIMIZATION IN PCM, ADAPTIVE PCM AND DPCM SYSTEMS , 1969 .

[4]  Andrew J. Viterbi.  Lower Bounds on Maximum Signal-to-Noise Ratios for Digital Communication Over the Gaussian Channel , 1964 .

[5]  J. M. Barstow,et al.  The measurement of random video interference to monochrome and color television pictures , 1962, Transactions of the American Institute of Electrical Engineers, Part I: Communication and Electronics.

[6]  R. Bruce.  Optimum Pre-Emphasis and De-Emphasis Networks for Transmission of Television by PCM , 1964 .

[7]  R. A. McDonald,et al.  Signal-to-noise and idle channel performance of differential pulse code modulation systems – particular applications to voice signals , 1966 .

[8]  Joel Max,et al.  Quantizing for minimum distortion , 1960, IRE Trans. Inf. Theory.

[9]  R. C. Brainard,et al.  Direct-feedback coders: Design and performance with television signals , 1969 .

[10]  P. Wintz,et al.  Quantizing for Noisy Channels , 1969 .

[11]  George C. Clark,et al.  Reconstruction error in waveform transmission (Corresp.) , 1967, IEEE Trans. Inf. Theory.

[12]  B. Cramer.  Optimum linear filtering of analog signals in noisy channels , 1966 .

[13]  Robert W. Donaldson,et al.  Optimum Pre- and Postfiltering of Sampled Signals with Application to Pulse Modulation and Data. Compression Systems P- , 1971 .

[14]  P. Wintz,et al.  The Effects of Channel Errors in DPCM Systems and Comparison with PCM Systems , 1973, IEEE Trans. Commun..

[15]  P. Noll,et al.  Effects of channel errors on the signal-to-noise performance of speech-encoding systems , 1975, The Bell System Technical Journal.

[16]  Colmas Goldberg.  Information Rates of Gaussian Signals Under Criteria Constraining the Error Spectrum , 1964 .

[17]  John P. Costas.  Coding with Linear Systems , 1952, Proceedings of the IRE.

[18]  J. Makhoul,et al.  Linear prediction: A tutorial review , 1975, Proceedings of the IEEE.

[19]  H. Spang,et al.  Reduction of Quantizing Noise by Use of Feedback , 1962 .

[20]  Daniel Esteban,et al.  Low‐bit‐rate voice transmission based on transversal block coding , 1976 .

[21]  Boris Tsybakov,et al.  Information transmission with additional noise , 1962, IRE Trans. Inf. Theory.

[22]  E. G. Kimme,et al.  Synthesis of optimal filters for a feedback quantization system , 1963 .

[23]  Ke-Yen Chang,et al.  Analysis, Optimization, and Sensitivity Study of Differential PCM Systems Operating on Noisy Communication Channels , 1972, IEEE Trans. Commun..

[24]  John B. O'Neal.  Bounds on subjective performance measures for source encoding systems , 1971, IEEE Trans. Inf. Theory.

[25]  P. R. Drouilhet,et al.  Asymptotically optimum pre-emphasis and de-emphasis networks for sampling and quantizing , 1966 .