Ancestor based survivor decision in M-algorithm convolutional decoding

In this paper, we propose a new algorithm for the surviving path decision in M-algorithm convolutional decoders. Correct path loss introduces one of the most destructive effects on the M-algorithm which mostly leads to catastrophic error. In the proposed M-algorithm survivor decision scheme, the M surviving states are not selected solely based on their own path metric. Among the M surviving states, some states with the best path metrics survive and also some other states whose ancestors have had the best path metrics survive. This way of survivor selection enables us to avoid correct path loss if an abrupt noise deteriorates the correct path metric severely. Simulation results show that the error rate performance of the ancestor-based survivor decision in some cases is slightly better than currently-best survivor decision scheme in the presence of the additive white Gaussian noise (AWGN). However, it is expected that the proposed algorithm offer better performance than the conventional methods in the presence of abrupt noise (short-term high value noise) like shot noise or over fading channel