Abstract
The commonly used rule of thumb of 5 m for the truncation depth of a memory m convolutional code is accurate only for rate 1/2 codes and should be replaced by two to three times m/(1-r) for a rate r code.

This publication has 7 references indexed in Scilit: