Blind Source Separation Based on Non-Gaussianity

摘要

Independent component analysis is an efficient way to solve blind source separation, which has been broadly used in many fields, such as speech recognition, image processing, wireless communication system, biomedical signal processing etc. Independent component analysis for the traditional ways to solve the blind source separation problem only considers the non-Gaussian signal, without taking into account the time structure of the signal information. Proposed based on generalized self-related and non-Gaussian source separation method, the full account of the non-Gaussian signal and time structure information, to solve the blind source separation problem in the time structure of the signal. Finally, this simulation method is validated, the simulation results show that the method is effective and worthy of promotion.