Turbodecoding method with re-encoding of erroneous information and feedback

摘要

Turbodecoding method with re-encoding of erroneous information and feedback The invention in general terms relates to a method of improving turbodecoding by re-encoding erroneous information and subtracting their contribution at the input of the turbodecoder. The subtraction of this contribution remedies the lack of convergence or convergence towards erroneous solutions observed in certain turbodecoding configurations. The invention applies equally well to parallel concatenation turbodecoding, to serial concatenation turbodecoding or to block turbodecoding. Several embodiments are presented according to the type of feedback envisaged.