Encoding and decoding rate-1/n convolutional codes and their punctured versions

摘要

The present invention is directed to the encoding and decoding of a digital signal. The encoding process results in a rate-1/n convolutional code derived from a rate-1/2 convolutional code. The process includes: selecting a base convolutional encoding rate of rate- 1/l, where l is an integer; selecting an output encoding rate of 1/n, where n is an integer greater than 1; encoding an input digital signal into a convolutional code comprised of signals S(0) through S(l-1), the convolutional code having the rate 1/l convolutional code encoding rate; and providing a rate-1/n convolutional code, which is derived from the rate-1/l convolutional code, the rate-1/n convolutional code having N(i) copies of the rate-1/l signals S(i), where i is from 0 through 1-l and where the sum of N(i) is equal to n. The decoding process results in a digital signal estimated from received symbols which include rate-1/n convolutional code generated by the above encoding process and any noise that may have been introduced by a transmission medium. The process includes the step of: generating a signal pair from the received symbols, the signal pair having a first signal and a second signal which are suitable for decoding by a rate-1/2 convolutional decoder. The first signal is an average of a sum of encoded signals which correspond to positions in the rate-1/2 convolutional code encoded using a first generator polynomial. The second signal is an average of a sum of encoded signals which correspond to positions in the rate-1/n convolutional code encoded using a second generator polynomial. The method also includes a step of decoding the signal pairs using a rate- 1/2 convolutional decoder.