DPCM image transmission over noisy channel using sequence minimum mean squared error decoder

Nowadays, joint source channel coding has become an important research topic. In this paper, source decoder is designed which minimizes the mean squared error in the reconstruction of a linear autoregressive DPCM coded signal over a noisy channel. The proposed sequence MMSE is based on the sequence-based approximate MMSE with first order Markovian approximation algorithm. The method with the best performance is compared with the sequence MAP technique. It is found that the method outperforms sequence MAP technique.