In this correspondence we study the decoding problem in an uncertain noise
environment. If the receiver knows the noise probability density function (
pdf) at each time slot or its a priori probability the standard Viterbi alg
orithm (VA) or the a posteriori probability (APP) algorithm can achieve opt
imal performance. However, if the actual noise distribution differs From th
e noise model used to design the receiver, there can be significant perform
ance degradation due to the model mismatch. The minimax concept is used to
minimize the worst possible error performance over a family of possible cha
nnel noise pdf's. We show that the optimal robust scheme is difficult to de
rive; therefore, alternative, practically feasible, robust decoding schemes
are presented and implemented on VA decoder and two-way APP decoder. Perfo
rmance analysis and numerical results show our robust decoders have a perfo
rmance advantage over standard decoders in uncertain noise channels, with n
o or little computational overhead. Our robust decoding approach can also e
xplain why for turbo decoding overestimating the noise variance gives bette
r results than underestimating it.