Fast simulation of turbo codes over AWGN channels

Abstract

In this paper, we illustrate the use of importance sampling (IS) for efficient simulation of Turbo Codes over additive white Gaussian noise (AWGN) channels. In particular, we show how to bias the probability density function (pdf) of the AWGN process using the most probable error path (MPEP) method in order to maximize the computational efficiency gain. The new noise pdf is designed based on the optimal biased distribution where zero variance estimator can be achieved. It is demonstrated that the proposed method requires much smaller sample sizes to achieve the same accuracy required by a conventional MC estimator.

Efficient performance evaluation of Forward Error Correcting codes

“…In the high SNR region, the probability of error is dominated mainly by the distance properties of turbo codes which depend on the interleaver design between the two constituent encoders. For turbo codes, an IS method based on the distance properties and error patterns was presented in [5].…”