Robustness of data compression coding schemes for still pictures over noisy channels

Abstract

A comparative study of the performances of five data-compression coding schemes over noisy channels is presented. The schemes are adaptive cosine transform coding, DPCM (differential pulse code modulation), adaptive block coding, vector quantization coding, and progressive transmission coding. Transmission is considered both with and without error control. For the former case, relations among bit error rate, signal-to-noise ratio (SNR), and mean code length are discussed. The influence of bit errors in the control information for adaptive coding is also studied. For the latter case, two error control procedures are examined: automatic repeat request and forward error correction. Some tradeoffs between increase of the code length and improvement of SNR by the error control procedures are discussed.<>