Classification and distribution of extrinsic information in turbo decoding for future wireless communications

The performance of turbo code is improved by updating the extrinsic information in the iterative decoding process. The transition patterns are categorized by the characteristics of the extrinsic information in this paper. And the distribution of these patterns is surveyed according to signal-to-noise ratio. Based on the results, the dominant error pattern is determined at high signal-to-noise ratio. Thus, the improvement of error performance in the error floor region of turbo code is expected by correcting the dominant error pattern which is found in this paper.