List decoding of turbo codes is analyzed under the assumption of a max
immm-likelihood (ML) list decoder, It is shown that large asymptotic g
ains can be achieved on both the additive white Gaussian noise (AWGN)
and fully interleaved flat Rayleigh-fading channels, It is also shown
that the relative asymptotic gains for turbo codes are larger than tho
se for convolutional codes. Finally, a practical list decoding algorit
hm based on the list output Viterbi algorithm (LOVA) is proposed as ar
e approximation to the ML list decoder. Simulation results show that t
he proposed algorithm provides significant gains corroborating the ana
lytical results, The asymptotic gain manifests itself as a reduction i
n the bit-error rate (BER) and frame error rate (FER) floor of turbo c
odes.