Turbo Equalization in Symmetric Α-Stable Noise Environments

Mingguo Zhao

2010-01-01

Abstract:This paper presents a turbo equalization method for the additive symmetric α-stable(SαS) noise environments.The segmented-polynomial approximation of the SαS probability distribution function was used in the maximum-a-posterior soft information iterations to fully utilize the prior information of the non-Gaussian noise.Simulations show that the method is superior to the standard Gaussian turbo method and the truncated turbo method proposed by Chuah.In moderate impulsive noise environment(α=1.5),the signal noise rate(Eb/N0) of the method outperforms 2.2 dB,better than that of the truncated method with a bit error rate of 10-4,and about 0.8 dB far from the Eb/N0 limit in the convolution-coded AWGN channel.The results also indicate that the method has better convergence performance.