Performance analysis of the dual sign algorithm with contaminated-Gaussian noise

In previous analysis of the dual sign algorithm (DSA), Gaussian signals were assumed. Thus it is inadequate when the desired response is corrupted additively by impulsive interference. The contaminated-Gaussian (CG) density is usually used to express the impulsive interference. A performance analysis of the DSA is presented when the input signal is a zero-mean stationary Gaussian process and the additive noise of the desired response is a zero-mean stationary CG process. Computer simulations validate our analysis and furthermore show that the DSA is less vulnerable to impulsive interference than the least mean square (LMS) algorithm.<<ETX>>