A Novel BER Model for Turbo Coding over Rayleigh Channel

The sensitivity of channel bit error rate (BER) to the turbo coding parameters (i.e., code rate, iterative number and interleave length, etc) and the wireless channel statistic enlightens us to model a BER for turbo coding over Rayleigh channel. The model presented in this paper synthetically considers and quantificationally analyzes the impacts of all the coding parameters and the wireless channel statistic on the BER, and then emphasizes a simple equation that relates the BER at the encoder to the relevant coding parameters. Based on this model, the BER performance can be estimated even before channel coding given certain coding parameters set and channel statistic. The simulation results show that the model can handle the actual values with high accuracy. The maximal average predictive error is around 0.07%. Moreover, the proposed model is very regular and simple, and can be extended to other cases. These characteristics will make it more useful in practice.