Linear System Identification Employing Independent Component Analysis

This paper presents a new approach for linear system identification with additive noisy output. It treats the model of system identification as an independent component analysis (ICA) problem with source signals received by several observed signals so that the estimation of noise can be obtained from the observed signals. By using some special characters of the mixing matrix, the ambiguity inherent in ICA is settled, and then, the parameters of the unknown system are obtained. This proposed approach does not rely on any statistic characteristics of the additive noise and can work well under low SNR conditions. Synthetic data are applied to validate the effectiveness of the proposed method and improved performance is obtained.