Turbo decoding performance in non Gaussian noise channels

Most of results in channel coding theory have been accomplished basing on the additive white Gaussian noise (AWGN) channels assumption. This hypothesis is not fulfilled in realistic environments in communications applications. In this paper, we investigate the robustness of the turbo code decoding algorithm in presence of a generalized Gaussian (GG) noise. This noise model has been proposed for modeling in many applications such as audio and speech processing. The probability density function (PDF) of such model depends on a shape parameter which determines the decay rate of the distribution. This paper presents the decoding performances of turbo code in terms of bit error rate (BER) when the noise PDF departs from Gaussianity into a GG distribution.