Implementation issues in MAP joint source/channel coding

A MAP (maximum a posteriori) approach to joint source/channel coding has been proposed which uses a MAP decoder and a modification of the souce coder to provide error correction. The authors present various implementation strategies for this approach and provide results for an image coding application. It is shown that, while the previously proposed approach achieves the objective of increasing the error-correcting capability, it also results in a variable-rate system. For this situation the previously proposed branch metric becomes difficult to implement. The authors explain these difficulties and propose implementable approximations of the metrics in the next section. Simulation results are given which demonstrate the viability of these approximations. The use of a variable-rate coder also complicates the structure of the decoder. The authors present a modified Viterbi decoder which can be used with variable-rate codes.