112 Adaptive Impulsive Noise Cancellation

Since impulsive noise in the application may usually appear, a novel adaptive impulsive noise cancellation approach (referred as LMPANC) is proposed based on Alpha stable distribution. In this proposed method, the adaptive filter is adapted. The impulsive noise is input signal of filter, and the useful signal with impulsive noise is the reference signal of filter. Both impulsive noises are relative. The filter coefficients are iterated under the least mean p-norm criterion. Upon convergence, the impulsive noise in the useful signal is cancelled mostly. Computer simulation studies show that the estimation performance of the LMPANC algorithm is excellent under non-Gaussian Alpha stable distribution noises, better than that of the traditional least mean square adaptive noise cancellation (referred as LMSANC) algorithm.