Threshold Based Iteration Stopping Criterion for Turbo Codes and for Scheme Combining a Turbo Code and a Golden Space-Time Block Code

Abstract

This paper proposes an iteration stopping criterion for turbo decoding with Benedetto's decoding algorithm based on a posteriori probabilities. This stopping criterion is used in two schemes. Firstly, it is used in a classical turbo code scheme on additive white gaussian noise (AWGN) channel. Secondly, it is used in a scheme combining a turbo code and a Golden space-time block code on fast Rayleigh fading multiple input multiple output (MIMO) channel. Simulation results with different thresholds for the stopping criterion show that a threshold of 1.2 and 1.4 in the first and second scheme, respectively, are sufficient for obtaining the same bit error rate and frame error rate performance like in the case of using the ideal genie stopping criterion. The difference between the average number of iterations for these thresholds and for the genie stopping criterion is at most 1.5 and 1.25, respectively.

Keywords

• average number of iterations
• space-time block codes
• stopping criterion
• threshold
• turbo codes