Abstract:
Numerical error characteristics are studied at the output of level quantizers. The errors occur in rounding off with the use of deterministic correcting signals. Conditions are obtained under which the numerical characteristics are independent of probabilistic characteristics of the processes being quantized.