Sliding Window Algorithm for TBCC Decoding

This study modifies the wrap-around Viterbi algorithm (WAVA) for tailbiting convolutional codes (TBCCs). The proposed modification is based on the fact that it is unnecessary to decode a codeword of TBCCs from the beginning of the codeword. WAVA must apply the Viterbi algorithm (VA) to the whole codeword repeatedly until the optimal sufficient condition is satisfied. Instead of checking the whole codeword, the algorithm checks the sufficient condition immediately after performing each level of metric computation. This approach significantly reduces the computational complexity. Simulation results in the case of (2,1,6) codes and E_b/N_0=3.5 dB show that the average number of extra levels is only 0.48, which is less than 1/9 that of WAVA.