An average-entropy-value-based stopping criterion for BICM-ID

Abstract

Stopping criterion is an efficient technique to reduce the unnecessary iterations of the iterative decoding process. Up to now, almost all the existing stopping criteria are designed for turbo-decoder, not for bit-interleaved coded modulation with iterative decoding (BICM-ID) receiver. Since the differences between the iterative schemes in turbo-decoder and BICM-ID receiver, these existing stopping criteria need greater computational complexity or extra storage when they are used in BICM-ID. This paper proposes an efficient stopping criterion, called AEV criterion, for BICM-ID receiver. This proposed criterion is based on average entropy value (AEV) of the decoding results. It takes a better trade-off between the average iteration numbers and the performance degradation than the existing criteria at high SNRs for BICM-ID. Furthermore, it needs no extra storage. Simulation results show that at high SNRs, with the increasing of SNR, the average iteration number curve of AEV criterion is getting closer and closer to the one of GENIE criterion which is the ideal but not a practical criterion. By using the proposed AEV criterion, the average iteration number is obviously reduced and the performance degradation is very small.