A robust ICA-based adaptive filter algorithm for system identification using stochastic information gradient

This paper proposes a new adaptive filter algorithm for system identification by using an independent component analysis (ICA) technique, which separates the signal from noisy observation under the assumption that the signal and noise are independent. We first introduce an augmented state-space expression of the observed signal, representing the problem in terms of ICA. By using a nonparametric Parzen window density estimator and the stochastic information gradient, we derive an adaptive algorithm to separate the noise from the signal. The computational complexity of the proposed algorithm is compared with that of the standard NLMS algorithm. The local convergence is analyzed. Due to the additive noise is also on-line estimated during the iteration, the proposed algorithm shows excellent robustness. It can directly be applied to an acoustic echo canceller without any double-talk detector. Some simulation results are carried out to show the superiority of our ICA method to the conventional NLMS algorithm.