A new stopping criterion for iterative decoding

We propose a new stopping criterion for iterative decoding of concatenated codes. The new stopping rule observes the a priori probabilities of different symbols and stops further iterations in the decoder whenever the average number of symbols in a trellis section below a certain a priori value, is above a predefined threshold. According to numerical examples for some practical transmission systems, the proposed criterion saves more decoding iterations than existing stopping rules. Also, it performs very close to the maximum possible limit of any stopping rule, with a very small degradation in bit error rate (BER) performance.