Abstract
Error protection coding is often used to combat the effect of transmission errors in binary encoded speech, but, if data throughput is to be kept constant, it requires an increase in transmission bit rate. An alternative approach, which does not use a protection code, decodes the erroneous speech at the receiver, inspects it, deduces the effect of transmission errors, and partially corrects it using a simple statistical hypothesis

This publication has 0 references indexed in Scilit: