Near-maximum-likelihood decoding of convolutional codes

Describes two novel decoding processes that come close to achieving maximum-likelihood decoding of convolutional codes, without, however, requiring nearly as much computation per decoded data symbol as does the Viterbi-algorithm decoder. The particular application studied is that of a bandwidth-efficient data transmission system, transmitting a rate-2/3 convolutionally coded 8PSK signal over a linear channel in the presence of additive white Gaussian noise. Results of computer simulation tests are presented, comparing the tolerances to additive white Gaussian noise of the new decoders with that of a more conventional near-maximum likelihood decoder, and also with that of a Viterbi decoder.