A unified framework for SNR mismatch in turbo algorithms

In this paper, we propose a unified framework to examine the effect of SNR mismatch on the performance of turbo algorithms, such as turbo equalization and turbo decoding. The proposed framework is based on a general model for a biased estimate of the signal-to-noise ratio (SNR) of the channel. It is shown how the estimation model can unify previously SNR mismatch models, which become special cases. Finally, simulation results show the effect of SNR mismatch on the bit error rate (BER) performance of a turbo equalizer based on the biased estimate of the channel SNR.