Abstract
Fading channels impose an intrinsic parametric uncertainty as a threat to reliable communications. In the case perfect channel information is present, optimal maximum a posteriori (MAP) algorithms were investigated in the literature. This paper focuses on the case of unknown channel information. We propose a MAP algorithm which takes the parametric uncertainty of the fading phenomenon into account by marginalizing the related probability distributions over the fading levels. The algorithm is applied to a serially concatenated system that employs continuous phase modulation as the inner code and a turbo code as the outer code.