Abstract
Abstract
Online bit error rate (BER) estimation (OBE) has been used as a stopping iterative turbo decoding criterion. However, the stopping criteria only work at high signal-to-noise ratios (SNRs), and fail to have early termination at low SNRs, which contributes to an additional iteration number and an increase in computational complexity. The failure of the stopping criteria is caused by the unsuitable BER threshold, which is obtained by estimating the expected BER performance at high SNRs, and this threshold does not indicate the correct termination according to convergence and non-convergence outputs (CNCO). Hence, in this paper, the threshold computation based on the BER of CNCO is proposed for an OBE stopping criterion (OBEsc). From the results, OBEsc is capable of terminating early in a varying SNR environment. The optimum number of iterations achieved by the OBEsc allows huge savings in decoding iteration number and decreasing the delay of turbo iterative decoding.
Subject
Electrical and Electronic Engineering
Reference48 articles.
1. Early stopping turbo decoders: A high-throughput, low-energy bit-level approach and implementation;IET Commun.,2010
2. A simple stopping criterion for turbo decoding;IEEE Commun. Lett.,2000
3. On the new stopping criteria of iterative turbo decoding by using decoding threshold;IEEE Trans. Signal Process.,2007
4. Log-likelihood values and Monte Carlo simulation-some fundamental results;Proc. Int. Symposium on Turbo Codes and Related Topics,2000
5. Threshold determining for minabsLLR stopping criterion for turbo codes;Frequenz,2013
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献