Iterative decoding of serially concatenated convolutional codes applying the SOVA

In numerous researches, soft output Viterbi algorithm (SOVA) has been used for the iterative decoding of a parallel concatenated convolutional code (PCCC) instead of the maximum a posteriori (MAP) decoding algorithm having a suboptimal bounds to the maximum likelihood (ML) performance. In this paper the authors propose a modified SOVA for the iterative decoding of a serially concatenated convolutional code (SCCC). The most important reason for using the SOVA is that it can achieve real-time decoding without noticeable information losses. Another advantage is that the SOVA requires less structural modification of the conventional decoder than other iterative decoding algorithms such as maximum a posteriori (MAP). The simulation results show that the proposed SCCC decoding scheme using the SOVA is superior to the noniterative SCCC. However, when compared with the scheme using the MAP algorithm [6], the performance gain is lower, as expected.