A SYSTEM APPROACH TO QUANTIZATION AND TRANSMISSION ERROR

In a system designed to quantize the output of an analog data source and to transmit this information over a digital channel, errors are introduced by the quantization and transmission processes. Quantization resolution can be improved by using all positions available in a data stream to carry information, or transmission accuracy can be improved if some of the positions are used for redundancy with error‐correcting codes. The problem is to determine, from a system viewpoint, the proper allocation of the available positions in order to reduce the average system error rather than concentrate exclusively on either the quantization problem or the transmission problem. We develop a criterion for the performance of data transmission systems based upon the numerical error that occurs between the analog source and the destination. The criterion, termed the average system error, is used to evaluate and compare possible system configurations. Significant‐bit packed codes are defined. These codes are useful because their protection can be matched to the numerical significance of the data and their redundancy can be sufficiently small to maintain good quantization resolution. The average system error resulting from representative system designs is numerically evaluated and compared. © 1969 The Bell System Technical Journal