Blind Channel Equalization in Impulse Noise

This paper presents a robust blind channel equalization technique in impulse noise environments. To suppress impulse noise effectively, the combination of order statistics (OS) and adaptive thresholding is considered. We implement nonlinear adaptation with the OS operation based on the Sato error criterion. The adaptive threshold is calculated using the variance of the input signal. Computer simulations demonstrate that a significant improvement in both the bit error rate and mean square error is achieved for communication channel models with a mixture of additive white Gaussian noise and impulse noise, in which two different impulse noise generation models with different noise generation probability are considered. KeywordsBlind channel equalization, Impulse noise, Order statistics, Threshold