Blind Source Separation Algorithm Using Nonparametric Adaptive Kernel Estimator

Nonparametric adaptive kernel density estimator (NAKDE) based blind source separation(BSS) algorithm is proposed under the framework of natural gradient optimization method. In order to improve the performance of source signal separation by BSS method, the probability distribution functions of source signals must be described as accurately as possible. Compared to the nonparametric fixed-width kernel density estimator(NFKDE) method, the NAKDE can improve the performance. Moreover, the direct estimation of the score functions can separate the hybrid mixtures of sources that contain hybrid both symmetric and asymmetric distribution source signals and do not need to assume the parametric nonlinear functions. The effectiveness of the proposed algorithm has been confirmed by simulations.